On The Internet, Nobody Knows You're A Dog. At The Supreme Court, Nobody Cares.

62 Comments

Kilroy • Mar 4, 2013 @2:35 pm

Don't tell lies to your wife and don't keep drugs in your car and you have nothing to worry about.
Bill • Mar 4, 2013 @2:44 pm

Can I keep drugs in my wife while she lies in the car?
Ken • Mar 4, 2013 @2:44 pm

Epic. Good to see you.
Clark • Mar 4, 2013 @2:45 pm

+1 point for a good post

+20,00000 points for embedding a classic Zevon tune.

(sorry; this is the internet)
Tali McPike • Mar 4, 2013 @2:48 pm

My brother had an run in with drug dogs once. My family had just moved to a small farm town about an hour outside of Bismark, ND. For some reason, the town was not very welcoming of outsiders, and with the exception of a few families, no one liked my brother and sister, including a lot of the adults.
I don't remember the exact circumstances around the incident, as this was over 2 years ago now, but a cop insisted on having a drug dog inspect the car he was driving (our mom's minivan). The unfortunate thing was that my mom had left a entire case of hemp milk in the back seat (about half my family has allergies of various sorts, and hemp milk is the only dairy alternative that no one was allergic to). Between that and the handler, its no wonder the dog gave positive feedback. Thankfully he wasn't arrested. But there was a nasty rumor about him doing drugs that ran around the town for a while.

And after another indecent with the local police improperly handling a break in at the house, my family decided to leave the town for good.
Hayden • Mar 4, 2013 @2:57 pm

Patrick, why do you keep a bull in your front yard?
Patrick • Mar 4, 2013 @2:57 pm

Minivans are very popular drug-smuggling conveyances, Tali McPike. My dog told me so!
machintelligence • Mar 4, 2013 @3:17 pm

But the dogs are certified (by the people that train them) so they must be accurate, right? [/snark]
Grandy • Mar 4, 2013 @3:25 pm

This is the price of ensuring that [Our Side] maintains control of [That Hot Topic Issue] in the supreme court. Anyway, eternal vigilance isn't needed when the state is doing your thinking for you.
Caleb • Mar 4, 2013 @3:26 pm

@ Tali

You have to remember that these are just simple farmers…
naught_for_naught • Mar 4, 2013 @3:35 pm

For some reason, the town was not very welcoming of outsiders.
Canadians have ruined it for everyone else, what with them buying up cheap Costco milk intended for god-fearing, bulk-shopping Americans.

[youtube http://www.youtube.com/watch?v=bOR38552MJA&w=560&h=315

Nothing to do with drug dogs, I know.
adam • Mar 4, 2013 @3:56 pm

man, thank goodness SCOTUS found in favor of Aldo, can you imagine the travesties that could/would occur if people found out we can't trust a dog's nose? (unanimous decision, my a$$)
Chris Berez • Mar 4, 2013 @4:14 pm

Really fantastic post, Patrick. I miss your regular posting, but man, when you do have something to say, you always smash it out of the park.
Shawn Young • Mar 4, 2013 @4:50 pm

I'm afraid I don't quite understand how it finished up: Patrick writes: "The truth is that last month your Supreme Court , in Florida v. Harris, bought that jive hook, line, and sinker." And yet he continues with "The truth is that the Florida Supreme Court got it right: the word of a dog is meaningless as the mathematics of Clever Hans…"

What am I missing here? (Thanks: my dog sniffed out a stash I had somehow hidden from myself, and I'm enjoying it, and…)
Patrick • Mar 4, 2013 @5:01 pm

We put up links because we want you to click them.
Maggie McNeill • Mar 4, 2013 @5:20 pm

This may give you some (very small) cause for optimism: last Thursday, in conversation with a group of about 6-7 law students, I brought up the dog case; every single one of them without exception thought the Exalted High Court was out of its collective gourd.
Christopher • Mar 4, 2013 @6:01 pm

Holy shit, this opinion.

So, first off, the opinion is, essentially, "Given that this is a good idea in theory, requiring any evidence that it actually works in practice would be unduly burdensome for the state."

That is a ludicrous opinion and it is galling that it was unanimous.

Furthermore, here's an actual quote from the opinion: "Conversely (and more relevant here), if the dog alerts to a car in which the officer finds no narcotics, the dog may not have made a mistake at all. The dog
may have detected substances that were too well hidden or
present in quantities too small for the officer to locate. Or the dog may have smelled the residual odor of drugs previously in the vehicle or on the driver’s person."

In other words, in actual field conditions, the dog may be so finely calibrated that it frequently indicates the presence of drugs in instances in which no drugs are to actually be found. These false positives shouldn't discourage us from treating the dog's alert as
probable cause because

The court also seems to think that because some minute trace of drugs are there, the dog hasn't, in these false positives, "made a mistake".

But (question from a non-lawyer) isn't the question whether the officer has a good reason to believe he will find evidence of a crime when he performs a search? If so, what does it matter why the dog gives false positives?
Tali McPike • Mar 4, 2013 @6:14 pm

@Caleb
You sir, made me snort my glass of koolade.
I haven't seen Blazing Saddles in a long time.
My dad is a native of ND (and grew up on a farm) and pretty much all of his family is still in various parts of the state, sadly its a fairly accurate representation.
Patrick • Mar 4, 2013 @6:17 pm

To answer your question Christopher, I'll ask you to define "good reason," which is essentially what the Supremes have announced, "logical" and "common-sense."

The Supreme Court has one opinion on what that means, that relying on an animal with a diploma is a logical, and common-sense, way to decide that a car likely contains drugs. Because the animal has a diploma.

I have another opinion. Like my friend Justice Kagan, I've studied Roman history. I know that the Emperor Caligula promised to name Incitatus, his horse, Consul before Caligula was assassinated.

That's evidently good enough for Justice Kagan. She would no doubt take a policeman's word, based on the word of Incitatus, who'd been vouched by no less an authority than the Emperor of Rome, that the car might contain drugs.

I'd say it's a fucking horse.
Kirk Taylor • Mar 4, 2013 @7:14 pm

Excellent Reason article on the subject: http://reason.com/archives/2013/01/31/this-dog-can-send-you-to-jail
Randy • Mar 4, 2013 @7:21 pm

Refuse a search in Texas you're gonna sit there till the dog comes while they threaten to impound. I've had it done to me twice. Twice the dog alerted and twice nothing was found because I simply don't do drugs nor associate with anyone that does. It's amazing how badly they screwed up this ruling.
Roscoe • Mar 4, 2013 @7:30 pm

Patrick – you suggest that if the good justices had tried a few cases they would be smarter and wouldn't reach such wrongheaded results. I don't know, I have tried quite a few cases and I am not sure I am any smarter for having done so.

The real problem is that these nine people, none of whom have any real experience at practical police work, have, through the auspices of the fourth and fifth amendments, set out to create a detailed set of rules for police procedure.

It didn't use to be this way. It was not until well into the 20th Century that the Supreme Court first started telling federal law enforcement agencies how to do their jobs. It was not until the 60s that the Supremes decided it was also their responsibility to tell state and local cops what they could and could not do. Now, of course, they are deciding when an elementary school vice principal can search a student's locker.

Way back in the day Judge Friendly wrote an interesting article called "The Bill of Rights as a Code of Criminal Procedure," suggesting that day-to-day rules of police procedure should be left to the people and their elected governments. Instead the Supreme Court has chosen to occupy the field, setting both a floor and a ceiling for police conduct.

The Supremes have no demonstrated body of knowledge of dog behavior. In spite of this there is nobody left to set the standards for when a dog alerting on something constitutes probable cause. And, when they get it wrong, well, they have lifetime tenure and are completely insulated from the political process.
James Pope • Mar 4, 2013 @7:34 pm

We've already got pieces of paper that are people, why not infallible dogs? We should just get the dogs into religion now, my dog likes funny hats.
Grandy • Mar 4, 2013 @7:59 pm

Roscoe, Patrick is suggesting that the backgrounds of our judges are too similar not that trying a few more cases would help. On the contrary, what we need are not simply judges with practical experience but practical criminal defense experience (and probably ideally, some time on both sides of that aisle).

The Supremes are a diverse lot on one level. But not so much on another; the one that matters when it comes to crime and punishment in this country. Indeed, they're frequently authoritarian on this subject. This decision was unanimous, and yet the science to support this is junk, and inarguably so.

It suits Washington just fine, of course.
Christopher • Mar 4, 2013 @9:02 pm

What amazes and baffles me is not just that they accept the diploma as a good reason to trust the dog, but that [i]they themselves[/i] explicitly lay out why the diploma may not make the dog competent at detecting evidence of a crime in the real world, [i]even if the testing is done competently[/i] and [i]the dog really can detect drugs[/i].

It's like, not only are they accepting senator horse, they're doing so after carefully explaining why a horse can't fulfill the duties of the senate.
P Mike • Mar 5, 2013 @7:08 am

Truly confused; this guy was not convicted on the basis of doggie testimony, just searched. There is no reason that the crtieria for searching should be the same as conviction in a court of law. (If it were, every time a search is made you could skip the trial and go straight to conviction.)

The question ought to be, "Is it reasonable to search if there is some evidence of drugs, even it if could be wrong?" If the answer is no, then anything goes (observing a drug deal? How did you KNOW it was a drug deal? Couldn't it have been anyting else? Couldn't your CI have been lying – after all he lied to the defendant?)
Patrick • Mar 5, 2013 @7:36 am

No, that isn't the question. The question before the Court was whether it is reasonable for an officer to search a car based on the behavior of one animal?

The Florida Supreme Court held that the animal's behavior did not create probable cause for a search, because the animal's track record (in only two cases, both involving Harris) showed that the animal would always behave in such a fashion as to indicate the presence of drugs, whether or not there were drugs to be found.

The United States Supreme Court held that the animal's behavior did create probable cause, because the animal has a diploma.

That's the range of potential answers. I leave it to you to decide which answer is correct.
P Mike • Mar 5, 2013 @7:47 am

"The Florida Supreme Court held that the animal's behavior did not create probable cause for a search, because the animal's track record (in only two cases, both involving Harris) showed that the animal would always behave in such a fashion as to indicate the presence of drugs, whether or not there were drugs to be found."

So was that the total track record?
Shawn Young • Mar 5, 2013 @8:09 am

Oh, I think I get the answer to my question. State SC went one way, SCOTUS reversed. Back to sleep with me then.
David Schwartz • Mar 5, 2013 @8:23 am

P Mike: We don't know this dog's total track record because police are careful never to test their dogs in ways that would give us reliable measures of how accurate they are. However, when they are tricked into testing their dogs in reliable ways, the dogs almost invariably alert to the places where their handlers think the drugs are.
En Passant • Mar 5, 2013 @8:26 am

After this brilliantly reasoned decision securing our rights against unreasonable searches, I expect them to articulate a new standard to secure our rights against compelled confessions, "If it floats it's a witch. If it drowns, it's not."

Without a wise Latina as bulwark against atavism, they would say, "If she floats, she's a witch. If she drowns, she's not."
Ken • Mar 5, 2013 @8:34 am

This is entirely consistent with SCOTUS' movement towards a anonymous-tipster policy that amounts to endorsing "a little bird told me."
tsrblke • Mar 5, 2013 @9:05 am

I submit that we're about to have "drug testing machines" replace dogs.
Of course those machines will have a button that allows them to alert on command (as a test feature I'm sure) which will cause them to always alert.
But based on this decision it's acceptable because the machines will be certified! Searches for everyone!
Grifter • Mar 5, 2013 @9:07 am

That dog in glasses is the cutest thing evar, btw.
En Passant • Mar 5, 2013 @9:13 am

Ken wrote Mar 5, 2013 @8:34 am:

… anonymous-tipster policy that amounts to endorsing "a little bird told me."

Well, that's a tried and true method for murdering innocent 92 year old grandmothers like Kathryn Johnston back in '06. So no doubt they'll find it furthers a compelling state interest.
Patrick • Mar 5, 2013 @9:20 am

The person who usually wears those glasses wishes he was remotely as cute.
A Different William • Mar 5, 2013 @10:12 am

Out of curiosity, what do you think is a reasonable false positive rate for probable cause?

If a drug dog indicates drugs when there are no drugs 5% of the time? 10%?
Dan Weber • Mar 5, 2013 @10:39 am

My inner geek would want to see an ROC curve for the dogs.

An "acceptable" false positive rate also depends on the circumstances where the dogs are used. If used on every Terry stop, that's different from going door-to-door.
James Pollock • Mar 5, 2013 @11:05 am

One of the problems with determining the accuracy of drug-sniffing dogs is that the only way to test is, well, with other drug-sniffing dogs.

You can build in an objective test where there's either a known presence of drugs or not, of course, but you can't tell whether or not the dog smells drugs, where the scent could have transferred.

I would be interested to know if dogs, once certified as drug-sniffing dogs, have to keep re-certifying, and if so, whether the dog re-certifies alone or with their usual handler.
Patrick • Mar 5, 2013 @11:26 am

You would have to conduct a blind study where the handler didn't know about the presence of drugs either.

I don't know much about drugs, but I know a lot about dogs. If I hide something from my dogs, they can generally figure out where it is by reading my face and my body language. At that task, they're better judges than most humans.
A Different William • Mar 5, 2013 @12:17 pm

Patrick, my family did that experiment with our dog. Our dog is trained to find money by smell. One person hides a dollar bill, then leaves the area. Then the handler and the dog come in to try and find the dollar bill.

The few times the dog did not find the bill he was in the right area and had it down to within 3 or 4 feet, but could not pin it down the rest of the way.

The experiment is fairly easy to do, and it is not hard to make it double blind.
princessartemis • Mar 5, 2013 @12:24 pm

Dogs will give humans whatever meaningless answers they want if it means praise and food and affection. They are highly tuned manipulators of humans; it's what we have bred them to be. That the Supreme Court didn't know this is pathetic.
Dan Weber • Mar 5, 2013 @12:34 pm

I'm sure dogs can find money and drugs.

These leaves some open questions:

1. What do they do when there aren't drugs or money to be found?

2. The dogs are not objective tools finding contraband, like a breathalyzer that gives a number. (I've never actually used one, so YMMV.) There are human trainers who are interpreting everything the dog does. It's as easy to game as someone doing cold readings.
A Different William • Mar 5, 2013 @1:00 pm

I'm sure Police officers can find money and drugs.

These leaves some open questions:

1. What do they do when there aren't drugs or money to be found?

The dog is a tool. Just like the breathalyzer. It gives a binary response: Bark or no Bark. Just like with the breathalyzer it is up to the cop to do his/her job objectively. In the breathalyzer case, if the value is above X then the cop takes you to the drunk tank. In the dog case, if the response is drugs are present the cop searches your car for them.
Dan Weber • Mar 5, 2013 @1:09 pm

But it's not "bark" or "no bark." The reason-dot-com article linked above has this nice quote:

Properly trained police dogs are supposed to indicate the presence of drugs with a clear, objectively verifiable signal, such as sitting down in front of an odor’s source or scratching at it. Yet “the dog never sat down, the dog never scratched, the dog never did anything that would indicate to me that it thought there was something in there.”
Grandy • Mar 5, 2013 @1:33 pm

Different William, we don't need to rely on your "test". As the Reason article notes, research suggests wildly varying degrees of accuracy with regards to the ability of dogs to sniff drugs. None of the levels of accuracy is what any sane person would call acceptable.
A Different William • Mar 5, 2013 @2:15 pm

I agree. my 'test' is anecdotal evidence. I am not arguing that drug sniffing dogs are always correct, but that dogs have the ability to smell drugs.

Field performance is a different matter. That depends on the officers using and maintaining their tools. Poorly trained officers and dogs will result in a very high false positive rate. However that does not mean that all drug dogs and officers who use them have that high of a rate.

I also did not see any citations in the Reason article. Do you know where the original studies are? I do not have time to look for them now, but I may go hunting later.
James Pollock • Mar 5, 2013 @2:43 pm

"You would have to conduct a blind study where the handler didn't know about the presence of drugs either."

My point is that even if whatever drugs the trainers are using are removed, the scent of drugs may either still be there or may have arrived from a different source.

So the results may be:
1) drugs are there, and the dog finds them.
2) drugs are not there, and the dog so indicates
3) drugs are not there, and the dog false positives
4) drugs are not there, but the scent of drugs is, and the dog detects that.

There's no way to tell 3 and 4 apart, unless your sense of smell is as acute as a dog's is.

I'm not disputing that A) the dog can be wrong, and B) the dog reads cues from the handler. The question is, how do you build in a restraint that works to reduce false positives without limiting accurate positives? Better training is the only one I can think of. Of course, in theory an agency doesn't want it's officers tied up searching people and places that don't have drugs. But does that agency interest carry down to the officers in the field?
MathMage • Mar 5, 2013 @3:20 pm

Patrick, my family did that experiment with our dog. Our dog is trained to find money by smell. One person hides a dollar bill, then leaves the area. Then the handler and the dog come in to try and find the dollar bill.

The few times the dog did not find the bill he was in the right area and had it down to within 3 or 4 feet, but could not pin it down the rest of the way.

The experiment is fairly easy to do, and it is not hard to make it double blind.

Did different people hide the bill each time? Was the bill's hiding place appropriately randomized? Did the person hiding the bill take care to rub their scent all over everything else so you could be sure the dog was sniffing for the bill and not for where Bill had been? Did you do controls where (a) nobody entered the room to hide a bill, and (b) somebody entered the room, went through all the motions of hiding a bill, then removed the bill? Did the dog see the person who hid the bill at any point between when he hid the bill and when the dog sniffed for the bill? Did the person guiding the dog know whether a bill had been hidden or not?

From the information you've given, it sounds like your anecdotal test is missing a large quantity of rigorous precautions.
MathMage • Mar 5, 2013 @3:28 pm

So the results may be:
1) drugs are there, and the dog finds them.
2) drugs are not there, and the dog so indicates
3) drugs are not there, and the dog false positives
4) drugs are not there, but the scent of drugs is, and the dog detects that.

Also:
5) Drugs are there, but the dog does not find them (perhaps because the handler wasn't expecting the dog to find them)

But that's sheer nitpickery; your main point is sound.
Joe Pullen • Mar 5, 2013 @5:45 pm

How long before they walk the dog up to your front door, have them "alert" on your house, and use that as a means to obtain a warrant for entry or perhaps to enter without a warrant citing possible spoilation of evidence.
ShelbyC • Mar 5, 2013 @6:30 pm

@Joe Pullen, How long? Negative a few years? You mean the other Florida drug-sniffing dog case before the court this term?
Dan Weber • Mar 5, 2013 @6:57 pm

How long before they walk the dog up to your front door, have them "alert" on your house

It's not often I can quote "Beetle Bailey" in these threads, but one Sunday strip had Otto growling at Beetle, and Sarge said it was because dogs can sense when something is wrong. By the last panel Beetle is in the stockade, and Killer asks him what he's in for, and he sadly responds "bad vibes."
Joe Pullen • Mar 5, 2013 @7:03 pm

@Shelby @Dan – Yeah that was somewhat rhetorical question. I like to pretend the 4th Amendment still exists and that the Supreme Court is doing the right thing. I’m into denial lately –it takes the edge off of reality.
sev • Mar 5, 2013 @9:38 pm

Wow, it's disturbing how many people have absolutely no idea what the problem is here — or how to test a drug dog's effectiveness in the real world.

The problem isn't necessarily that the dogs have a false positive rate, it's that in all cases the handler can generally induce false positives, even if the handler isn't aware of it. And if they do it properly, the false positive rate is 100%. Therefore, bringing a drug dog into the picture means that the officers will -always- have 'probable cause' for a search if they want to make the dog fake a positive — or even if they're just prejudiced in some way against the subject, and in this case, if they think that there's probably drugs in the car or on the subject, they're prejudiced.

It's relatively easy to test the efficacy of drug dogs. You do a double-blind test with boxes and drug samples.

However, to test the efficacy of the dog/handler system, you have to lie to the handler. Tell them that there're no drugs, but you want them to run through their paces anyway, when there are indeed drugs on the course. Tell them that there're drugs in the course when there aren't.

That's what everyone forgets. A dog is not a computer. It's not a binary system looking at one input. The handler is as much of a part of the 'instrument' here as the dog is.

The combination of honest error plus accidental, prejudicial biasing plus deliberate biasing means that drug dog/handler teams, as they are used now, are an extraordinarily crappy way to get probable cause. Though I'm sure that many officers are above the board and as honest as possible when using their dogs, I'm just as sure many are not. Get stopped in Texas looking like a scruffy long-haired hippy freak, and you can be -sure- that the drug dog, if they call one in, will alert.
En Passant • Mar 5, 2013 @10:15 pm

sev wrote Mar 5, 2013 @9:38 pm:

The problem isn't necessarily that the dogs have a false positive rate, it's that in all cases the handler can generally induce false positives, even if the handler isn't aware of it.

Cf. Handler beliefs affect scent detection dog outcomes, Lit, Schweitzer and Oberbauer, Anim Cogn. 2011 May; 14(3): 387–394.
James Pollock • Mar 6, 2013 @12:06 am

"Also:
5) Drugs are there, but the dog does not find them (perhaps because the handler wasn't expecting the dog to find them)
But that's sheer nitpickery; your main point is sound."

I wasn't attempting to present an exhaustive list, just enough to highlight 3 and 4. For example, there's also
6) dog bites handler and
7) dog passes gas so noxious it clears the room.
wgering • Mar 6, 2013 @1:47 am

I would totally pay $5,000.00 for a dog that could deliver rousing speeches about scalping Nazis.
Andromachos • Mar 6, 2013 @5:05 am

I guess a solution would be to have one's own dog trained for drug sniffing. After the police dog false alerts run your own dog around for no alert. Voila no probable cause.

Well, if the cop doesn't shoot your dog. Alerts often look like "aggressive behavior"
Brian Utterback • Mar 6, 2013 @8:51 am

Perhaps I am missing something, but from the opinion it looks like the SCOTUS didn't say that the dog was reliable, but instead said that the officer had reason to believe that the dog was reliable. They specifically said that the reliability of the dog could be challenged. I know this is a subtle difference, and brings with it its own problems (it disincentivizes the police from keeping reliability records on the dogs), but it isn't the same as buying the reliability of Aldo "hook, line and sinker".
Jonathan Kamens • Mar 8, 2013 @5:39 pm

It astounds me that do many people here, including Patrick, either didn't bother to read the decision, didn't understand it, or are intentionally misrepresenting it.
The court did not rule that the dog's alert established probable cause because the dog has a diploma. The court ruled that the prosecution presented evidence of the dog's reliability AND THE DEFENSE FAILED TO CHALLENGE IT AT TRIAL. That was the main issue under consideration by the court and indeed the only issue that should have been considered on appeal. I'm not a lawyer, and even I know that you don't get to introduce new evidence in appeal unless it was not available during trial.
"… the Court does not consider such arguments when they are presented for this first time in this Court."
Sometime claimed in a comment that the problem is that SCOTUS is trying to tell police specifically how to do their jobs. Did the person who made this comment actually read the decision? Because what the actually say in the decision is that courts should NOT be imposing strict checklist tests on how police do their jobs, and that the Florida Supreme court erred by doing so.
This is not a civil rights issue. This is an issue of a defendant having incompetent counsel who failed to aggressively challenge the evidence against him at trial, and incompetent appeal lawyers who failed to appeal I the grounds of incompetent counsel at trial, an appeal they might have actually won.
Malc • Apr 1, 2013 @12:42 pm

Thank you, @Jonathan Kamens! This was NOT a 4th Amendment case, not a case about the fallibility of dogs, not a case about randomized double-blind tests.

This was simply a case about how and when evidence to suppress should be introduced. The answer is… "Not on appeal".

I think no biscuit for Patrick…