Wednesday, March 28, 2012

SCOTUS to consider what training, credentials needed for drug-dog alert probable cause

The US Supreme Court agreed this week to hear arguments regarding a case from the Florida Supreme Court to decide what level of training, certification, etc., is sufficient for a narcotics dog's alert to count as probable cause. The Florida Supreme Court held the dog in Florida v. Harris was not sufficiently credentialed, so if SCOTUS doesn't like the ruling, there's a risk they took it up to overturn it.

The opinion out of Florida is fascinating, and they appear to have caught the state in a Catch-22: "The State argues that records of field performance are meaningless because dogs do not distinguish between residual odors and drugs that are present and, thus, alerts in the field without contraband having been found are merely unverified alerts, not false alerts. This assertion, if correct, raises its own set of concerns as it relates to a probable cause determination of whether the dog's alert indicates a fair probability that there are drugs presently inside the vehicle."

The Florida court declined "to adopt the view of the First, Fourth, and Fifth Districts" because it would "place the burden on the defendant to uncover all records and evidence that might challenge a presumption of reliability — evidence that is exclusively within the control of law enforcement authorities and, further, evidence that law enforcement agencies may choose not to record, such as in this case." Absent such documentation, "when a dog alerts, the fact that the dog has been trained and certified is simply not enough to establish probable cause to search the interior of the vehicle and the person."

Bottom line, if I'm reading the opinion correctly, the dog in Harris had been "trained" and "certified," but the department kept no ongoing training records and considered their hit-rate in the field meaningless and so didn't record it. Thus the court ruled that merely being trained and certified, at some point in time by somebody, is not enough to judge a dog reliable without more documentation. "Because the State must establish that the officer has a reasonable basis for believing that his or her dog is reliable in order to prove probable cause based on the dog's alert," said the Flordia court, "the State carries the burden of presenting the necessary records and evidence for the trial court to consider in adequately evaluating the dog's reliability."

How many K-9 handlers keep training and field records recording accuracy rates, and what proportion of US drug dogs might this case affect if SCOTUS accepts the Florida Supreme Court's view? Who knows? Or maybe they just took the case up to bench slap the Florida high court and bring it in line with the more permissive federal districts?

This is an area of forensics - used on the front end for probable cause as opposed to the back end at trial - that deserves much more scrutiny. Dog alerts in other contexts have been deemed unreliable. As the court noted, "there is no uniform standard in this state or nationwide for an acceptable level of training, testing, or certification for drug-detection dogs."

I'm hardly sanguine the US Supreme Court will agree with their brethren jurists from the Sunshine State, but Grits welcomes the debate. Even if they strike down the Florida ruling, the case will be an opportunity for a belated discussion over what dog handlers should be doing with regards to training and record keeping compared to what often happens in the field.

MORE: See an article Radley Balko wrote last year on the reliability of drug dog alerts and how humans interpret them.

AND MORE: See a discussion of drug dogs on the Texas prosecutor association user forum in which Williamson County DA John Bradley confidently predicts that "SCOTUS took the case to remind the states that probable cause is not cause for a full-blown trial." Tarrant County prosecutor David Curl, though, noted that the Florida court emphasized that "evidence of the dog's performance history in the field — and the significance of any incidents where the dog alerted without contraband being found — is part of a court's evaluation of the dog's reliability under a totality of the circumstances analysis. In particular, when assessing the factors bearing on the dog's reliability, it is important to include, as part of a complete evaluation, how often the dog has alerted in the field without illegal contraband having been found."

FURTHER THOUGHTS: We don't know which justices wanted to hear the case, so it's hard to say before oral arguments whether Harris will be more about reinforcing the court's decision in Caballes giving carte blanche for drug sniffing dogs, or more about policing drug dogs' misuse in light of a greater awareness of forensic errors. I"m not a lawyer, but as it's been explained to me, "probable cause" generally means "more likely than not," or above a 50% likelihood. ("Reasonable suspicion" is an even lower standard.) So if a drug dog finds contraband only 45% of the time it alerts - to pull a number out of the air - would that constitute "probable cause," or must the dog's reliability be higher? And what happens when a law enforcement agency avoids that question simply by failing to keep records? Those are the questions Grits hopes the court will focus on when it takes up Florida v. Harris.


Anonymous said...

I read the state's motion for cert. They try to say that anytime a dog alerts and drugs are not found it is attributable to residual odor. That is simply not the case. Sometimes the dogs just falsely alert and, very often, a false alert is due to handler cuing, often unintentional. Handler cuing is also referred to as the Clever Hans effect. Clever Hans was a horse that could count. Actually the horse was responding to miniscule cues from his handler.

In some research I did, I discovered that the actual accuracy rate of these dogs, if used at random on the general population, is probably between 15% and 20%. The only reason we see higher accuracy rates is because they are not used at random on the general population. They are typically deployed in situations where the officer already has developed some suspicion. This significantly increases their accuracy rates. This supports the Florida Supreme Courts position that the courts should look at the totality of the circumstances and probable cause should not be based on a dog's alert alone. Requiring the officers to develop additional reasonable suspicion isn't a big deal because, in many cases, they are already doing that, they just don't realize it.

But, back to the main issue of field records...the reason they are not kept by some departments is the plain and simple fact that they show the dogs aren't perfect. The idea that every false alert is attributable to residual odor is just ridiculous. Unfortunately, many, many courts have adopted this ridiculous logic.

As far as certification goes.. I think it was Radley Balko who wrote a good article about the fact that there is an association that certifies psychics. So, using the logic as the courts have applied it to dogs, a certified psychic could be enough to support probable cause.

The idea of certification has some issues that are similar to those in the discussion about crime labs. The certification organizations are run by people with ties to law enforcement, law enforcement is their biggest customer, and if use of the dog's fell out of favor these organizations would cease to exist. So, there is an inherent bias in these organizations which leads them to inflate the accuracy of these dogs.

Anonymous said...

4:25 here again.

There was an interesting study of these dogs done by UC Davis. This study was done in a church, so the false alerts cannot be attributed to residual odor:

The study, published in the January issue of the journal Animal Cognition, found that detection-dog/handler teams erroneously “alerted,” or identified a scent, when there was no scent present more than 200 times — particularly when the handler believed that there was scent present.

The dog-handler teams conducted two separate five-minute searches of each room. When handlers believed their dogs had alerted — indicated a target scent — an observer recorded the location indicated by handlers. Search orders were counterbalanced; that is, all teams searched the rooms in a different order.

Although there should have been no alerts in any of the rooms, there were alerts in all rooms. Moreover, there were more alerts at the locations indicated by construction paper than at either of the locations containing just the decoy scents or at any other locations.

Anonymous said...

4:25 here again.

Another issue with these dogs is the so called "passive" alerts. A dog is trained to alert in a certain way. Yet, handlers will often testify that a long sniff or something like that was an alert where that is not what the dog was trained to do. To be blunt, these handlers are simply lying so that officers can search the car. Courts need to require that there be an objectively observable alert in the manner in which the dog was trained.

Anonymous said...

Anon 4;25 et al,

You mentioned some research you did regarding actual accuracy rates. Was it written up somewhere and is it accessable? Thanks.

Gritsforbreakfast said...

FWIW, 8:53, according to Radley Balko, "Several studies and tests have shown that drug-sniffing dogs, scent hounds, and even explosive-detecting dogs are not nearly as accurate as they have been portrayed in court. A recent Chicago Tribune survey of traffic stops by suburban police departments from 2007 to 2009, for example, found that searches turned up contraband in just 44 percent of the cases where police dogs alerted to the presence of narcotics. (An alert is a signal, such as barking or sitting, that dogs are trained to display when they detect the target scent.) In stops involving Hispanic drivers, the dogs' success rate was just 27 percent."

One good thing about SCOTUS hearing this case is we'll know a lot more about these questions with data and studies from "both" sides before it's all done. Not much attention is normally paid to the issue by either the public nor independent researchers.

Anonymous said...

It would be interesting to compare the training and documentary procedures used by bomb dogs v. drug dogs. It is my understanding that the type of documention the court discussed is actually recorded with bomb dogs.

As an aside, years ago I transported about 400 kilos of high quality cocaine in the trunk of a car for a short distance. About 12 to 16 hours later, I had a goveernment drug dog check out the trunk of the sedan, even opening the trunk for the dog, yet it never hit on the auto, bringing into question the whole issue of residual odor for me.

Anonymous said...

8:53, when I said I did some research I meant that I looked at studies and papers that had been written on the subject. I did not do any studies myself. Here are the sources I found:

This is the Radley Balko article -
Radley Balko, The Mind of a Police Dog, REASON.COM (February 21, 2011),

The UC Davis study noted above is probably the most interesting -

Probably the best analysis I found was in this article - Richard E. Myers II, Detector Dogs and Probable Cause, 14 GEO. MASON L.REV. 1 (2006).

Here's a good quote from the Myers article:
With a pretty good dog, but a largely innocent population, a dog alert will signal drugs only about sixteen percent of the time. The reason is this: Because the officer is stopping mostly innocent people, one has to be more concerned about the false positive error (alerting when there are no drugs). Because there are more cars without drugs in them, the gross number of searches that result from the error rate will be higher than the gross number of searches that result from correct alerts. Overall, there will be many more searches of innocent people than there will be searches of guilty people.

Here's a good article that applies some good statistical analysis: Robert C. Bird, An Examination of the Training and Reliability of the Narcotics Detection Dog, 85 KY. L.J. 405, 420 (1997).

Also, in just reviewing a lot of court cases, it appeared that accuracy rates were typically in the 40% to 60% range, but remember that they are only this high typically because officers usually have developed some suspicion before deploying the dogs. Often when you see a handler asserting a higher rate you will find that they either don't record false-positives or they attribute them to residual odor.

Anonymous said...

I find it outrageous that our constitutional rights rest on the nose of a dog, no matter how well trained it is.

Anonymous said...

Grits - love the blog but IMO you've have a deeply flawed understanding of probable cause.

The standard you articulate i.e. "more likely than not" is a preponderence of evidence, the standard to find someone liable in a civil action.

Probable cause is definitively a much lower burden - most courts accept the definition as a "fair probability." SCOTUS pretty clearly rejected better-than-even odds in Illinois v. Gates.

Gritsforbreakfast said...

@1:00 - As mentioned, I'm not a lawyer, but certainly I know civil standards are different, the distinctions between standards are fuzzy, and anyone who tries to put percentages on those standards is generally on shaky ground.

Even so, it's hard to imagine how "probable cause" could be had in a situation where the pivotal indicator - e.g., a dog sniff that hypothetically "hits" 40% of the time - by definition means it's more likely than not that an innocent person is the target of an "alert." I.e., when a dog alerts, if they have a 40% "hit rate," in any given circumstance there's a 60% chance they find no contraband.

The dictionary defition of "probable" is "likely to occur or prove true," with the second definition, "having more evidence for or against."

To this non-lawyer, it seems like an Orwellian interpretation to make "probable" a synonym for "improbable." That doesn't mean you're wrong about the case law. According to Wikipedia, SCOTUS defined it as a "fair probability," which courts have interpreted as meaning "30%, others 40%, others 51%." If the latter, that's the same as the civil standard of "preponderance," which is how I've always understood it. I've been wrong before, though, and could well be again.

Anonymous said...

Here's an interesting quote from SCOTUS comparing the level of probable cause needed to search generally, and the standard required to search school children:

Perhaps the best that can be said generally about the required knowledge component of probable cause for a law enforcement officer's evidence search is that it raise a “fair probability,” Gates, 462 U.S., at 238, 103 S.Ct. 2317, or a “substantial chance,” id., at 244, n. 13, 103 S.Ct. 2317, of discovering evidence of criminal activity. The lesser standard for school searches could as readily be described as a moderate chance of finding evidence of wrongdoing.Safford Unified School Dist. No. 1 v. Redding
557 U.S. 364, 129 S.Ct. 2633

So, the term "fair probability" is equated with a "substantial chance" which is more than a "moderate chance." It seems that, in the case of dog sniffs, some want to use a standard lower than a "moderate chance." In my humble opinion a "substantial chance" should be at least 50% or more.

Anonymous said...

Anon 4:25 et al (and Grits)


Checkpoint USA said...

For those of you who may not already be aware, in some parts of the country, the use of drug sniffing dogs against the traveling public has become a ubiquitous practice. Specifically, Customs & Border Protection has outfitted nearly all their internal immigration checkpoints with K9 units. These drug dogs are routinely used against domestic traffic at pre-primary, primary & secondary inspection stations all absent any individualized suspicion in the vast majority of cases.

In discussions on the topic of the use of drug dogs during traffic stops, I hardly ever see reference to just how far down the road we’ve already traveled as a society. Usually discussions revolve around the use of drug dogs during traffic stops in which an officer claims either reasonable suspicion or probable cause existed to initiate the stop in the first place but few recognize the fact that the federal government has already made major in-roads to the use of drug dogs during traffic stops absent any suspicion whatsoever.

As such, hundreds of thousands of people who live and/or work within 100 miles of an international border, are subjected to daily suspicionless intrusions in their lives. I happen to be one of those individuals and began documenting my encounters years ago.

As evidence, I’ve compiled several links below to videos I’ve made & articles I’ve written about this subject on my blog, Roadblock Revelations. I’ve also included links to several other youtube videos I’ve come across on the subject. Finally, I have several dozen additional videos of drug dogs in action I’ve taken over the past few years that I’ve yet to process and post online. I’d be more than happy to release this data to anyone who thinks they can use it to help put a stop to these daily violations of our fundamental rights.

Links to videos & articles on the use of drug dogs at suspicionless checkpoints inside the country appear below:

* Illegal Homeland Security Drug Checkpoints:

* Give Them An Inch and They’ll Take 100 Miles:

* The Worst Part of Censorship Is ————-

* Customs K9 Team Deployed At Internal Immigration Checkpoint:

* Immigration Status Smelling Dog or Illegal Drug Checkpoint?:

* DHS Checkpoint Blog Entry 10: Man’s Best Friend:

* DHS Checkpoint Blog Entry 11: The French Connection:

* A Deja Vu Moment. Anderson Found Not Guilty – Again:

* Border Patrol Agent says dog NEVER makes a mistake!!!:

A few additional videos on youtube related to the use of drug dogs at internal suspicionless immigration checkpoints (these are not my videos):

* I-8 Border Patrol Drug Checkpoint:

* Border Patrol (100 miles from the border) :

* Civil Rights Violated by Border Patrol Agents in Arizona:

* Border Dog:

Anonymous said...

Perhaps a tea made from stems and sprayed on every car you pass,especially every police car,could give them plenty of search and fail statistics.