Skip navigation
PYHS - Header
× You have 2 more free articles available this month. Subscribe today.

Drug Detection Dogs Are Unreliable and Reflect the Vicious Heritage of Their Slave-Hunting Dog and Police-Dog Predecessors

by Matt Clarke
LAW ENFORCEMENT OFFICIALS OFTEN deploy trained dogs to detect drugs, but how accurate are the canine sniffs? Since the dog cannot testify, the courts have simply accepted that a certified drug dog is sufficient to provide probable cause for a search when the dog alerts its handlers, allegedly indicating the presence of illegal drugs. The failings of this legal fiction include a lack of standards for training, testing reliability, and certification, as well as research indicating that drug detection dogs frequently “alert” because of intentional or unintentional cues from their handlers.
When the problem of cuing a response was first documented in a 2011 study showing an 85% false-positive rate, handlers refused to cooperate with researchers attempting to find ways to eliminate unintentional cuing. But why would handlers want dogs to alert when no drugs are present? One possible answer may be found in law enforcement’s use of civil asset forfeiture (“CAF”) to seize assets suspected of being used for or gained through drug deals. Using a drug dog alert to establish probable cause to seize money or property for purposes of CAF, law enforcement agencies at all levels profit from false drug detection dog alerts.
Early History of Working Dogs
THE OLDEST PROOF OF A SPECIAL RELAtionship between humans and dogs is a 14,000-year-old late Paleolithic co-burial site in Germany. Scholars found evidence of dogs being used to assist humans hunting wild game in Japan 10,000 years ago. The military forces of ancient Greece, Rome, and Egypt are also known to have employed dogs in warfare.
After Europeans “discovered” the New World in 1492, they employed dogs against the Native population. There are numerous accounts of Christopher Columbus and other conquistadors using dogs that devoured human prey—including women and children—to terrorize Native populations into submission.
These war dogs, often given spiked collars and battle armor, were used in violent land grabs across the Western Hemisphere, as well as during the Reconquista attacks on the “Moors” (Muslims, Arabs, and North African Berbers) who were driven out of Spain during the same period. One account written in 1553 about the colonizer Francisco Pizarro, known for brutalizing Native people of Peru, said that his dogs were so fierce “that in two bites with their cruel teeth they laid open their victims to the entrails.” These war dogs are early examples of canines being trained and used in a clearly racist fashion.
Selective Breeding of Slave Dogs and Police Dogs
AS EARLY AS THE 16TH CENTURY, HUMANS began selectively breeding dogs for specific characteristics. The Cuban bloodhound became infamous as a vicious hunter of escaped slaves that mauled its victims, sometimes fatally. This served the enslavers’ purpose as an escape-prone slave was worth little, but a mauled slave was less likely to be physically capable of attempting an escape and served as a constant visible reminder to other slaves of the consequences of such an attempt. This breed of bloodhound arose around the time of the American Revolution and, during the first half of the 19th century, was extensively imported into the U.S. for use by individual enslavers and official slave patrols—antecedents of the current police.
Intentional selective breeding of dogs for police use began in Germany in the 19th century. The first school designed to train dogs for police work was founded in 1895 in Ghent, Belgium. Experiments conducted in 1896 proved the German Shepherd was a superior breed for police duties. By 1900, German, Dutch, and Belgian police departments were deploying trained German Shepherds.
The use of dogs by modern police forces in the U.S. came about in 1907, when New York City imported trained German Shepherds. By 1911, the city owned 16 dogs that were used to patrol Long Island.
Racialized Use of Police Dogs in the Civil Rights Era
WIDESPREAD CREATION OF K-9 PROGRAMS in U.S. police departments did not come about until the Civil Rights Era. The city of Baltimore created a “Canine Corps” in 1956, followed by St. Louis in 1958. They were likely inspired by a 1955 article in Police Science touting New York City’s successful use of police dogs. Soon police
departments throughout the country were creating K-9 corps and often deploying them against civil rights demonstrators. Once again, “law enforcement” was unleashing dogs against Black people.
One infamous incident occurred in Jackson, Mississippi in 1961 when two police dogs, Happy and Rebel, attacked Black college students during a sit-in at a segregated public library. The following day, the dogs were set loose on a large gathering of peaceful demonstrators outside the city courthouse. A local Jackson newspaper, the State Times, described the incident with the headline “Police Dogs Emphasize Law, Order” and enthusiastically noted that the dogs’ trainer was Harry Naworth, “the former Nazi storm trooper who trained killer Dobermans to guard Hitler’s airports.” According to the article, Naworth had trained “more than 23,000 dogs of all breeds during his lifetime.”
A defining moment in police dog use came in 1963 in Birmingham, Alabama, when police dogs were set loose on peaceful demonstrators as they entered the business district. The specter of “disorderly Black mobs” was used to justify vicious dog attacks—even against the elderly, children, and pregnant women. The head of a local NAACP at the time noted that “dogs basically aren’t used on white people.” After the Birmingham incident, Martin Luther King, Jr. famously said “we’re going on in spite of the dogs and fire hoses.” Malcolm X presciently noted that “100 years ago they used to put on a white sheet and use a bloodhound against Negroes. Today, they’ve taken off the white sheet and put on police uniforms. They’ve traded in the bloodhounds for police dogs and they are still doing the same thing.”
Early Police Dog Training
INITIALLY, POLICE DOGS WERE TRAINED TO “bite and hold” when unleashed to chase down a fleeing suspect. They were supposed to bite the person on an appendage and hold until the handler gave an order to release the hold. However, the dogs were also trained to escalate the violence—to tear and rend should the bitten person resist or attack the dog. The two problems with this were that dogs would “mistake” the upper thigh or the neck for appendages, initiating potentially fatal bites, as it is impossible to expect a person who is being bitten by a dog to simply remain motionless. Another issue was that police dogs who “got too far ahead” of their handlers had been known to attack innocent bystanders. Bad publicity and several large-award lawsuits led some trainers to preface “bite-and-hold” with another step, “bark and circle.”
A dog trained to bark and circle first tries to immobilize a person by circling the person while barking and only graduates to “bite-and-hold” or “tear-and-rend” if the person continues to flee or attacks the dog. However, most police dogs are still trained with the “bite-and-hold” method.
Since the controversial, racist use of police dogs in the Civil Rights Era, police dogs and their handlers have once again been rebranded. Modern police dogs are thought of as detection devices used to sniff out drugs, explosives, or bodies. But there is no fundamental difference between a drug detection dog and a police dog of the Civil Rights Era, except that it has received additional training in drug detection. Drug detection dogs are still trained to attack people on command, protect their handlers, and bite-and-hold or bark-and-circle if unleashed to chase a fleeing person.
Given the brutal history of the dogs, it isn’t hard to see why people of color are nervous around police dogs. Yet, that pervasive fear is used against minorities to justify searches and forfeiture of money and other property, even when evidence of criminal activity is lacking. This “dog-induced nervousness” takes on an even greater role when one considers just how poor many detection dogs are at actually finding drugs.
First Study of Handler Effects on Detection Dogs’ Reliability
IN JANUARY 2011, A FORMER DETECTION dogs handler named Lisa Lit was a postdoctoral fellow in the department of neurology at the University of California, Davis and the lead author of a study published in the journal Animal Cognition. She later referred to the study as “a real career-ender.” Entitled “Handler beliefs affect scent detection dog outcomes,” the study showed that in 123 of 144 test runs (85%), detection dogs alerted when their handlers had been falsely told drugs or explosives were hidden in the area to be searched. The greatest number of alerts occurred at places that were marked, so handlers would “know” a stash was there. There were a total of 225 false alerts, with some runs having up to five false alerts despite the total absence of drugs or explosives.
The study, which was double-blind (meaning any information that may influence the tester or the subject is withheld until after the test), involved 18 dog handlers and their scent dogs. All of the teams had been certified by law enforcement—13 for drug detection, three for explosives detection, and two for both. The runs were performed in four rooms of a church, each of which was between 30 and 40 square meters, and the dogs searched all four rooms in a randomly-selected sequence for a total maximum time of five minutes. In two of the testing conditions, handlers were falsely told that the locations of the scent items were marked with a sheet of red paper. There were four possible conditions: control, paper marker, unmarked presence of a scent decoy consisting of two Vienna sausages and a tennis ball (to test the dog’s scent interest), and paper marker at location of the scent decoy. Each team was tested with each condition twice for a total of 144 runs. There were no explosives or drugs present at any time.
The result of the study, which surprised the researchers, was at least one false alert during 85% (123) of the runs. Only five teams achieved no-alert runs and only one had no alerts during all eight runs. The overall number of alerts was about the same for each condition—between 55 and 57—but the location most alerted on was the marker without a decoy scent (33 alerts). The marked decoy scent had 29 alerts, and the unmarked decoy scent had 18. All other alerts were at 36 other locations that were neither marked nor contained a decoy scent.
The study attributed the false alerts to the “Clever Hans” effect, referring to a horse that, in the early 20th century, was thought to be capable of counting and other “human” mental tasks until psychologist Oskar Pfungst showed the horse was responding to unintentional cues given by its trainer’s posture and facial expression. Lit’s intention was to quantify any problems associated with intentional or unintentional handler cuing and develop methods to solve the problem. However, after the study was published, handlers refused to cooperate in further investigations.
Other Studies of Detection Dog Reliability
IN JANUARY 2011, THE CHICAGO TRIBUNE analyzed three years of data from suburban police departments to determine the reliability of drug detection dog alerts performed on vehicles stopped by police. It found that only 44% of searches performed pursuant to those alerts turned up drugs or drug paraphernalia. For Hispanic drivers, the success rate was 27%. This means that the false alert rate was 56% overall and 73% for Hispanics, calling into question both the reliability of drug detection dogs and the racial bias in the stop-and-sniff system.
Handlers and their organizations often argue that it is impossible for dogs to have a racial bias. The history of training dogs to attack dark-skinned people shows that assertion to be disingenuous. More importantly, if dogs are taking cues from their trainers, they may be reflecting the trainers’ biases.
The Tribune data appear to support this likelihood in that some of the police departments had astonishingly low success rates with alerts on vehicles driven by Hispanics. For instance, in Naperville, the rate was 8%, compared with the combined average of 47% for all of the searches. Likewise, in McHenry County, the overall success rate was 32%, but it was a mere 12% for Hispanic drivers.
An Australian academic study published in Drug and Alcohol Review in 2012 determined that “no drugs were found in approximately three-quarters of police searches following a positive notification by a sniffer dog during a two-year period.” The study included interviews with drug users who attended music festivals where drug dogs had been employed. They reported that drug sniffing dogs were incapable of detecting drugs hidden in body cavities, they were generally undeterred from their drug use by the dogs, but some users engaged in the risky behaviors of purchasing drugs at the venue in response to the use of dogs, while others consumed all of their drugs, risking an overdose, when they spotted dogs working the crowds.
A second Australian study, published by RMIT University in 2018, surveyed attendees at large outdoor music festivals and reported a drug detection rate of 3% for the sniffer dogs. It noted a review conducted by the New South Wales Ombudsman office in 2006 that found a 74% false positive rate there and that only 1% of the positive indications ever led to a prosecution.
A 2014 study by Polish officials published in Forensic Science International discovered a great deal of variance in the abilities of different breeds of dogs to detect drug scents. It found German Shepherds to be the best at
drug detection and terriers to be the worst. It found alert accuracy rates of 83.2% for indoor searches, 63.5% for outside of cars, and 57.9% for inside of cars. The study also found that the scent of hashish can linger and be detectable for at least 48 hours. However, the study was criticized for not being double-blind or having conditions with no drugs present.
The ‘Legal Fiction’ Surrounding Drug Detection Dogs
“LEGAL FICTION” IS THE TERM USED TO describe an assumption and acceptance of something as a fact by a court, despite the fact it may not be, in fact, true in order to allow a rule to operate or be applied in a way that differs from its original purpose, leaving the letter of the law undisturbed. It is a judicial form of alternate reality to achieve a desired goal such as convenience, consistency, equity, or justice. One such legal fiction about police dogs was created in Robinette v. Barnes, 854 F.2d 909 (6th Cir. 1988), where the court concluded that, “the use of a properly trained police dog to seize a felony suspect does not constitute deadly force” even if “in this particular case the use of a police dog to apprehend a suspected felon resulted in that felon’s death.”
Another police dog legal fiction was implicitly created in United States v. Place, 462 U.S. 696 (1983), when the Supreme Court of the United States (“SCOTUS”) ruled that a sniff search by a dog is not a search requiring a warrant under the Fourth Amendment because of the narrow scope of the search and the information obtained, implying that detection dogs are infallible or at least extremely reliable in detecting the contraband in question.
In United States v. Trayer, 701 F.Supp. 250 (D.D.C. 1988), a federal district court ruled that a drug detection dog alert is sufficiently reliable to provide probable cause for a search. The U.S. Court of Appeals for the First Circuit built on this ruling in United States v. Owens, 167 F.3d 739 (1st Cir. 1999), when it ruled that an alert by a drug detection dog that had twice failed certification and had a false positive rate of over 50% was still sufficient probable cause for a search. In doing so, the court accepted the dog handler’s claim that dogs who alerted where no drugs were found were not giving a false positive, but rather alerting to the scent of drugs that had been previously present and removed prior to the sniff. This created a new legal fiction of residual odor causing false alerts.
In United States v. Outlaw, 134 F. Supp. 2d 807 (W.D. Tex. 2001), a federal court slammed the door on using a Daubert hearing to challenge the accuracy of drug detection dogs. See Daubert v. Dow Pharmaceuticals Inc., 509 U.S. 579 (1993) (standard used by courts to assess whether an expert witness’ scientific testimony is based upon scientifically valid reasoning that can properly applied to the facts in question).
The most important drug detection dog case is United States v. Caballes, 543 U.S. 405 (2005), in which SCOTUS held that the use of a well-trained drug detection dog on the exterior of a vehicle during a lawful traffic stop doesn’t implicate legitimate privacy concerns and is not a search for purposes of the Fourth Amendment. The Caballes Court reasoned that “[a]ny interest in possessing contraband cannot be deemed ‘legitimate,’ and thus, governmental conduct that only reveals the possession of contraband compromises no legitimate privacy interest.” In his dissent, Justice Souter wrote that the infallible drug detection dog is a creature of legal fiction, noting that a study used by Illinois in its brief showed that drug detection dogs failed tests in the lab 12.5% to 60% of the time, leading him to question the accuracy of drug dog alerts. Nonetheless, Caballes became the most cited and written about drug dog case and the foundation for future U.S. Supreme Court decisions.
Caballes involved a driver pulled over for speeding. A drug dog sniffed outside his car and alerted. The car was searched, and police found marijuana in the trunk. The Court called the dog sniff sui generis (in a class of its own) because it only revealed the presence of contraband. Thus, adopting the reasoning of the Place Court, the Caballes Court affirmed the concept of the infallible dog sniff and the lack of a need for a warrant to conduct a search following a dog alert. Inexplicably, it never addressed the issue of false alerts.
The defense in United States v. Funds in the Amount of Thirty Thousand Six Hundred Seventy Dollars, 403 F.3d 448 (7th Cir. 2005), tried to require the prosecution to prove a dog’s sniff is reliable. The appellate court ruled that the prosecution need not show the statistical reliability of the detection dog because it would be too complicated and the dog handler was better equipped than to court to determine how reliable a dog is. It
ruled that proof of certification is sufficient evidence of reliability. Clearly, this is yet another judicially created legal fiction.
SCOTUS doubled down on its infallible drug dog concept in Florida v. Harris, 568 U.S. 237 (2013), a case in which a drug dog alerted during a traffic stop, and pseudoephedrine was found, despite the dog not having been trained to detect pseudoephedrine. The Florida Supreme Court had ruled that a drug dog’s reliability had to be proven before it could be accepted as probable cause for a search. SCOTUS disagreed, holding that the dog alert was sufficient to provide probable cause as long as the dog is certified, but the Court failed to provide any guidelines for certification.
A question by Justice Scalia during oral arguments in Harris showed how disingenuous or naive he was regarding drug dog searches: “Why would a police department want to use an incompetent dog?” he asked the public defender. “What incentive is there for a police department?”
There’s at least one simple answer, “Follow the money!” Law enforcement uses civil asset forfeiture to seize assets suspected of being obtained by or used in criminal drug activities. Often, a drug dog alert is given as the probable cause for that allegation. Further, why require search warrants at all if we can be so trusting of the police? What is the difference between a search based on an unreliable drug dog alert and one based on a policeman’s hunch—the very thing the Fourth Amendment was drafted to prevent?
Two years after the Harris decision, United States v. Bentley, F.3d (7th Cir. 2015), made it even more difficult to challenge drug dog reliability. The dog handler in Bentley admitted that his dog alerted 93% of the time, yet drugs were found in only 59% of searches. The handler admitted that he only rewarded the dog when it alerted, possibly the reason for such a large 41% false positive rate. Noting that it had previously rejected a challenge to a dog with a false positive rate in excess of 60%, the Bentley Court ruled that, pursuant to Harris, certification is sufficient to establish drug detection dog reliability.
In 2018, the Court of Appeals of Maryland held that a trial court’s determination of a drug dog’s reliability is a question of fact to be reviewed for clear error on appeal. Grimm v. State, 183 A.3d 167 (2018). The trial court had admitted a great deal of evidence about the reliability of Ace, the drug detection dog in question. This included Ace’s training records, field reports, certification guidelines for drug detection dogs, and Ace and his handler’s certifications as well as testimony from the handler, the defense, and prosecution experts. The trial court found Ace was reliable.
The appellate court noted that determinations of the sufficiency of probable cause are reviewed de novo, but the background facts necessary to establish probable cause are reviewed for clear error under the totality of the circumstances. Drug dog reliability is such a background fact, and the Grimm Court found no clear error, affirming the trial court, but nonetheless opening up the possibility of future challenges to drug detection dog reliability.
In 2018, a divided Minnesota Supreme Court held that a drug dog sniff in the shared hallways of an apartment building is not a search requiring a warrant. State v. Edstrom, 916 N.W.2d 512 (Minn. 2018). SCOTUS had previously held that a dog sniffing on the porch of a house is a search requiring a warrant. Florida v. Jardines, 569 U.S. 1 (2013). Thus, Minnesota is extending lesser Fourth Amendment protections against warrantless searches to apartment dwellers, who tend to be poorer than house dwellers.
In a 2020 decision with great potential impact, the U.S. District Court for the District of Utah called into question the reliability of every drug dog in Utah. See United States v. Jordan, 455 F. Supp. 3d 1247 (D. Utah). The court examined the training of a drug detection dog named “Tank” who had been trained under the Utah Peace Officer Standards and Training (“POST”) guidelines. It astutely noted that, under POST, the handler can “call literally any movement of the dog an alert, turning normal dog behavior into probable cause for a search. It doesn’t help that the dogs are rewarded for every alert and given no positive reinforcement for failing to find anything interesting” when no contraband is present.
The court also noted that Tank was trained in only 28 exercises in the six months around the time of the search in question, all but one of which had drugs present. Further, the court found that Utah POST does not use double blind training for Tank and other drug detection dogs. It does not even use single-blind training, as
the handler always knows the number of drug stashes and often their location as well. This allowed Tank to cue off of his handler or the person observing the tests.
The court also viewed the video recording of the traffic stop and dog sniff, concluding that the handler “relied on Tank’s natural behaviors, which he perceived as ‘alerts,’ to conclude that Tank had detected the odor of narcotics emitting from” the vehicle and there was “nothing on the video of the sniff from which a third person can objectively conclude that Tank had performed to respond as he had been trained to do when he detected a target odor . . .”
“Behavior by the dog that is so subjective that only the handler may be able to identify it risks allowing a search in violation of the Fourth Amendment that is based on nothing more certain than the officer’s hunch that drugs may be present,” the court chided.
It is interesting that the handler was supposed to wear a body camera, but he did not. The video came from a detective who was on the scene. The view of the dog was at times blocked by vehicles, but the court refused to accept the handler’s testimony about what happened when the dog was off camera.
The court found that “Utah POST Training inadequately addresses, and therefore fails to remove the risk of, inadvertent handler bias or cuing. Specifically, Utah POST’s failure to implement double-blind training raises questions as to the independence of its K-9s and casts doubt as to whether the K-9s are alerting or indicating because they actually detect the odor of narcotics or because they have learned that displaying such action is the best way to please their masters. This doubt is not allayed by Utah POST’s certification process, as the final test that a K-9 must pass in order to be certified is not even single-blind.”
The court granted the defense’s motion to suppress the pipe-residue of marijuana and pistol discovered in the vehicle and, on the Government’s own motion, dismissed the indictment for lack of evidence. Jordan.
Despite the promise of Jordan and its possible implications for every drug-sniffing dog in Utah, it is still a decision by a single federal district court that is not binding precedent outside of that district or even on other judges in the same district. See United States v. Articles of Drug Consisting of 203 Paper Bags, 818 F.2d 569 (7th Cir. 1987). The current law of the land, as determined by SCOTUS in Caballes and Harris, is that proof that a drug detection dog is trained and certified constitutes proof that it is a reliable law enforcement tool, regardless of the actual facts, i.e., a legal fiction. Thus, it remains somewhere between difficult and impossible for defendants to challenge the reliability of drug detection dogs in court, despite the fact the science doesn’t support the reliability of drug detection dogs.
Drug Dogs’ Media Presence
DRUG DOGS ARE OFTEN GLAMORIZED IN THE media as being responsible for the seizure of enormous amounts of drugs and cash. It is estimated that the various police departments currently have over 600,000 dogs performing a variety of duties. In 2004, drug detection dogs were credited with the seizure of between $2 billion and $3 billion annually.
Dandy, a drug-sniffing German Shepherd in southern California, was credited with the seizure of over $1 billion in drugs and cash over his six-year career. Twelve drug detection dogs at a U.S. Border Patrol Station in El Paso, Texas, made possible the seizure of $100 million in narcotics over a nine-month period between 1988 and 1989.
Rarely mentioned in the media are the numerous times an alert by a drug detection dog leads to nothing—or, even worse, some poor soul’s insurance settlement or house down-payment being seized by police for simply having the misfortune of possessing a large amount of cash and encountering police officers on a fishing expedition. Also rarely mentioned in the media are dogs like Karma who signaled an alert to every single vehicle he sniffed during roadside traffic stops in Republic, a small city in Washington state. Drugs were found only 29% of the time, meaning he gave a false alert 71% of the time.
Other rarely-mentioned drug detection dog disasters include Bono, a Virginia drug detection dog who gave false alerts 74% of the time; Sella, a Florida drug-sniffing dog that gave false alerts 53% of the time; and Lex, who did the same in Illinois more than 40% of the time. Even if nothing is seized, a false alert results in a traumatic, humiliating, and time-consuming waste of at least one innocent citizen’s time—more than half an hour on average.
Civil Forfeiture: Follow The Money
“POLICING FOR PROFIT,” A 2020 REPORT BY the Institute for Justice, a libertarian think tank, noted that local, state, and federal law enforcement agencies had taken for themselves over $68.8 billion in civil forfeiture assets since 2000. One can hardly complain when police seize money found alongside large quantities of drugs, but most of the time, that is not the case. Instead, many of the seizures are of cash found in vehicles following a drug dog alert when no drugs are found.
Sometimes, the owners of seized cash win it back in court, but doing so costs them months or years of time and thousands or even tens of thousands of dollars in attorney’s fees. If the amount seized is a few thousand dollars or less—staggering sums to many—the cost of litigation exceeds the possible gain, and there is always the chance that you could lose. Seizures are rarely contested simply because the system is engineered to make it extremely difficult for the average person to successfully challenge forfeitures without the help of an experienced attorney, but the majority of seizures are for amounts far too little to justify the expense of hiring an attorney. So faced with that reality, most victims simply resign themselves to never getting their money returned, even if they are never charged or even arrested in connection with the seizure.
Examples Of Unjust Civil Forfeiture
REAL ESTATE AGENT AND GRANDMOTHER Wendy Farris was travelling from Montana to her grandson’s birthday party in Oregon when she became drowsy and pulled off the road at a safe spot to nap in Republic, Washington. A deputy sheriff found her there asleep behind the wheel, woke her, and ordered her to submit to a field sobriety test. She does not drink or use drugs as a blood test later confirmed. Nevertheless, the deputy arrested her for suspicion of driving while intoxicated and called a K-9 unit to the scene.
The infamous Karma showed up and, just as she had done at every single other drug sniff, alerted. That gave the deputy probable cause to impound and search the car, a Prius. No drugs or alcohol were found, but $4,656 in cash was discovered.
Instead of releasing Farris, the county held her without charges over the weekend and kept the cash and car. She eventually got both back but missed her grandson’s birthday, incurred legal bills, and was presented with a bill for the hygiene supplies she used while incarcerated in the jail.
Travelling from his son’s home in Philadelphia, Jerome Chennault was pulled over for allegedly “following too closely” in Madison County, Illinois. He had with him $22,870 for a down payment on a home in Las Vegas, Nevada. A drug dog alerted, and police found the cash—but, of course, no drugs. Chennault had to hire a local attorney and make repeated trips to Illinois from his home in Las Vegas. He eventually got his cash back, but his legal expenses were not reimbursed. Nonetheless, Chennault is one of the lucky ones.
Following a drug dog alert on Straughn Gorman’s motor home after two coordinated pretextual traffic stops in Nevada, the $167,000 he was taking to California to pay for a house was seized. No drugs were found. A trial court ruled in his favor and chided the U.S. Attorney’s office for failing to disclose critical facts favorable to Gorman, but the U.S. Attorney appealed the ruling. Chennault had to pay a bond (10% of the cash’s value) and continue to fight the case and accumulate more legal bills .
In January 2009, police in Putnam County, Indiana, pulled Anthony Smelley over for an alleged unlawful lane change and an obscured license plate. A drug dog alerted on his car, and $17,320 was found and seized. No drugs were found. Smelley said he was carrying the money to purchase a car for his aunt. He even produced a letter from a Detroit law firm stating that the money came from a settlement in a lawsuit over a recent car accident. Incredibly, the county moved forward with forfeiture proceedings on the bizarre theory that the money could have been used to purchase drugs sometime in the future.
A report by Reason showed that the county was being represented by a private attorney, Christopher Gambill, who specializes in civil asset forfeiture for the county and was paid a contingency fee of 25% to 33% of forfeited assets. After Smelley’s money was held for 13 months, a circuit court judge ruled that a drug dog alert, without any other evidence, is insufficient to provide probable cause for forfeiture. Lucky for Smelley, no officer claimed he was “nervous” or acting strangely, as other courts have held that a drug dog alert in conjunction with “abnormal” behavior can provide probable cause for forfeiture.
Civil Forfeiture’s Disproportionate Effect on Minorities and The Poor
THE CASE OF RUDY RAMIREZ, A TEXAS resident who was pulled over in Kansas City in 2000, is typical of most civil asset forfeiture incidents. A dog alerted, and the $7,300 in cash he was carrying to purchase an automobile was found. Police took $6,000 despite there being no evidence of drugs or other criminal activity. Lawyers told Ramirez it would cost $10,000 in legal fees to get the money back, so he let the cops keep it.
This is typical of people who are too poor to challenge the system of civil asset forfeiture or when the non-reimbursable costs of fighting the seizure exceeds the value of the seized property. The Washington Post found that there were 61,998 roadside seizures of cash totaling $2.5 billion between 2001 and 2014—an average of about $40,000 per seizure. Only a sixth of them were challenged and only 44% of the challenges were successful. Studies have shown that police disproportionately target minorities and vehicles with out-of-state license plates using minor traffic offenses that a driver would likely not be conscious of, such as straying in a lane or changing lanes without signaling, as a pretext to stop the vehicle and have a dog sniff it.
The Future of Detection Dog Training
GIVEN THAT MANY DRUG DOG HANDLERS refer to their dogs as “probable cause on four legs,” civil asset forfeiture should perhaps be renamed civil furfeiture, but the question of whether drug detection dogs can be trained to ignore handler cues and biases remains. Experts on drug dog training confidently state that the dogs can be trained to ignore cues from the handler and other people in the area. They also assure us that the dogs can be trained to ignore handler biases and the scent of amounts of drugs that are too tiny to be detected as well as residual drug odors. This is important because, as early as 1991, reports showed that as much as 90% of the bank notes in circulation are tainted with cocaine. The American Chemical Society reported on an international study published in 2009, indicating that the U.S. and Canada are among the countries with the highest levels of cocaine-contaminated bills, with an average contamination rate between 85 and 90%, according to Yuegang Zuo, Ph.D., a study leader from the University of Massachusetts in Dartmouth. Thus, dogs may alert to the presence of cash based on cocaine contamination, not whether there are identifiable amounts of drugs nearby.
Even the 2011 study by Lit showed that one detection dog/handler team got it right every single time. That supports the claim that the problem is training, not the innate ability of dogs to detect scents. Why, then, do we still have problems with the issue of handler cues and biases and alerts attributed to undetectable amounts of drugs or residual scent? The answer is that police departments do not want their dogs trained to ignore cuing. They know the answer to Justice Scalia’s question—having a dog that will alert on cue lets the police use their “four-footed probable cause” to search anybody they want to, even if the search is based on a mere hunch or more troubling motives.
Other Issues Related to Drug Detection Dogs
THE WIDESPREAD LEGALIZATION OF MARIjuana has created problems in the world of drug detection dogs, as almost all of them are trained to detect more than one drug, and marijuana is both the easiest drug for them to detect and one of the drugs almost all of the dogs are trained to detect. Those facts are forcing police departments to give drug detection dogs an early retirement and to purchase new police dogs at a cost of $8,000 each, then pay between $12,000 and $15,000 for patrol and detection training.
In 2021, the Associated Press reported that Virginia state police were retiring 13 K-9s, while many smaller police departments and sheriff’s offices were retiring one or two dogs. Some departments are unable to afford the cost of obtaining and training new dogs and are disbanding their K-9 units.
The Quincy, Massachusetts police department shifted two dogs from drug detection to patrol work after marijuana was legalized in the state in 2016. They retired both dogs about 18 months later. Lt. Bob Gillan, who heads the department’s K-9 unit, said that drug traffickers quickly learned how to use the legalization of marijuana to raise doubts about detection dog alerts.
“Usually, when they’re delivering their illegal drugs, they will always have marijuana burning in the car,” Gillan said. “Any defense attorney worth his or her salt will say, ‘Well, your dog hit on a legal substance [the marijuana],’ (not the illegal drugs).”
Some of the marijuana-trained detection dogs have been shifted to being used in places where marijuana is still illegal, such as schools. That is not without controversy, as shown by the public backlash against frequent school sweeps in Aurora, Colorado. There, teachers and students told the school board that students had been traumatized, not just by the dogs, but also by handlers and school police who tackled a student and slammed him to the ground before leading him away in handcuffs. A few marijuana edibles were found in his backpack.
“Why do we bring drug dogs into classrooms?” teacher Brian Lindstrom asked. “Do you feel like the trauma created in our community, that already has historical trauma, was worth the five edibles found? Because I don’t think it was worth it.”
WHETHER IN SCENT LINEUPS OR DETECTION work, reports have shown that dog handlers, intentionally or not, give cues and exaggerate their dogs’ abilities. This approaches—or in the case of scent lineups, achieves—the status of “junk science.” Add to this the fact that dog handlers have been convicted for falsifying evidence and are resistant toward training their dogs to ignore handler cues and biases, and one can only conclude that drug detection dogs and their handlers need national training and certification standards independent of organizations that have a monetary interest in making sure the dog is certified. Currently, competing certifying organizations such as the National Narcotic Dog Association and the National Tactical Police Dog Association sell certifications based upon their own standards. As the Jordan case shows, even having statewide training standards does not mean that the training produces reliable drug dogs. Until this is corrected, the courts must stop accepting proof of training and certification as conclusive evidence that a detection dog is reliable. There’s a good, yet deeply troubling, reason police officers jokingly refer to drug detection dogs as “probable cause on four legs.”
Sources: Northern Illinois Law Review, Animal Cognition,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,

As a digital subscriber to Criminal Legal News, you can access full text and downloads for this and other premium content.

Subscribe today

Already a subscriber? Login



Federal Prison Handbook - Side
Advertise Here 4th Ad
Prison Phone Justice Campaign