dog weed sniff ratio

Drug-Sniffing Dogs in Schools Make Every Student a Suspect

Last month, a high school in southeastern Washington conducted a suspicionless drug search. Students were asked to leave their classroom so that a police officer with a “drug-detection dog” could check their backpacks for signs of drug possession. After the search, two students were singled out for a more invasive search and questioning. One had marijuana paraphernalia in his backpack; in the other, no signs of drugs or drug paraphernalia could be found. Good news for the second student—after the humiliating and anxiety-producing search was complete, he was permitted to go back to class.

We can all agree that drug use and abuse among high school students is a serious concern. And most would agree that high schools should take measures to foster a drug-free environment. So what is the problem with such searches that use “drug detection dogs” at school? Aren’t drug dogs just a tool to help the police target student drug suppliers? They won’t cause any harm to students who follow school rules, so no need for us to worry about them. Right?

Not exactly. There are a number of reasons we should be concerned about drug-sniffing dogs in schools, but one key problem is that they have not been very effective at targeting only drug possessors. Several studies have indicated that drug dogs are prone to false alerts, which then lead to unjustified searches. Records of drug-sniffing dogs in one Washington school district indicated that dogs were incorrect 85 percent of the times they alerted to a substance. A Chicago study of drug dogs used for roadside automobile searches shows a 56 percent error rate—increasing to 73 percent for Hispanic drivers. Even the most generous estimates suggest that drug dogs are reliable, at most, 70 percent of the time (and this figure takes into account the 26% of searches where no substances are actually found but the targeted person admits to prior drug contact).

Why such high false alert rates? While some people have expressed concern that drug-sniffing dogs are influenced by officers’ racial biases, there are more benign explanations for the dogs’ inherent unreliability. For example, many drug-dog-and-officer teams may be poorly trained, dogs may be rewarded for alerting even when no substances are found, and police officers, who are often under pressure to show that costly drug-detection programs are working, may give dogs accidental cues.

Further, there is little or no evidence to support claims that these programs deter drug use, reduce drug-related crime, or increase perceptions of public safety. This means that the searches are not only costly to individuals who experience embarrassment, humiliation, or anger when they are searched based on a false alert. They are also costly to school districts. Maintaining a drug-dog program can cost a district between $12,000 and $36,000 every year. Given the severe budget cuts hitting Washington public schools (which have led to the elimination of dropout-prevention assistance and other programs that aimed to reduce the achievement gap), the cost-effectiveness of programs that identify one or two students possessing trace amounts of marijuana at a handful of schools should be seriously questioned.

The drug dog programs also raise a serious constitutional concern. Do public school students give up all expectations of privacy when they enter school grounds? A number of court cases have made clear that they do not. In Kuehn v. Renton School District (1985), a precedent-setting ACLU case, the Washington Supreme Court held that it is unconstitutional for public schools to search large groups of students without individualized suspicion of each person searched. A blanket search that treats every student as a suspect violates the “privacy clause” of the Washington Constitution, which provides that “No person shall be disturbed in his private affairs, or his home invaded, without authority of law.”

More recently, the Washington Supreme Court unanimously found that a school district policy of suspicionless urine testing for students who participate in extracurricular athletics violates the Washington Constitution (York v. Wahkiakum, 2008). The cases make clear that our state constitution provides even broader protections for privacy than its federal counterpart.

In 2006, the Nine Mile Falls School District in eastern Washington recognized that its blanket use of drug-sniffing dogs may have similarly violated the “privacy clause.” It abandoned its drug dog program in response to a threatened lawsuit by the ACLU of Washington and the Center for Justice. We hope that other school districts in Washington will acknowledge that drug dog searches are a bad policy and do the same.

Last month, a high school in southeastern Washington conducted a suspicionless drug search. Students were asked to leave their classroom so that a police off

Are drug sniffer dogs incorrect 75 per cent of the time?

RMIT ABC Fact Check

Article share options

Share this on

  • Facebook
  • Twitter
  • LinkedIn

Send this by

  • Email
  • Messenger
  • Copy link
  • WhatsApp

The claim

In support of their plan to legalise pill testing of party drugs in Victoria, the Greens have taken aim at drug detection canines, declaring:

“Drug sniffer dogs are incorrect 75 per cent of the time.”

Is that correct? RMIT ABC Fact Check investigates.

The verdict

The Greens’ claim is close to the mark.

The source of the claim, a review conducted in NSW between 2002 and 2004, showed that in 74 per cent of cases where a sniffer dog indicated the presence of drugs on someone, no drugs were subsequently found.

More recent figures from NSW show no drugs were found in 63 per cent of cases after a sniffer dog indicated their presence; for South Australia it was 82 per cent.

While police suggest the dogs are almost always correct in picking up either the presence of drugs or residual traces of drugs, the stated purpose of sniffer dogs is to detect people in possession of drugs — not traces of drugs that may indicate previous use.

The source of the claim

The claim appeared on a Victorian Greens website advocating pill testing — a harm reduction service in which drug users can get information on the content of illegal drugs, allowing them to make informed choices.

A spokesman for the Victorian Greens told Fact Check in an email that the source of the claim was a 2006 NSW Ombudsman’s Review of the Police Powers (Drug Detection Dogs) Act 2001.

Over the Ombudsman review period between February 2002 and February 2004, police drug-detection dogs indicated the presence of drugs on someone 10,211 times, with “almost all of these indications resulting in a search”.

However, according to the report, “prohibited drugs were only located in 26 per cent of the searches following an indication”.

In other words, in 74 per cent of cases where an indication from a drug dog formed the reasonable suspicion necessary for conducting a search, no drugs were found.

Dogs with jobs

In Victoria, there is no specific legislation covering the use of drug detection dogs by law enforcement authorities, but an information sheet prepared by Victoria Police outlines the rationale for their deployment.

A spokesman for Victoria Police told Fact Check drug detection dogs were used as a tactical tool in the investigation of drug use and trafficking.

“Under the Drugs, Poisons and Controlled Substances Act, police have the power to search for illicit drugs in situations when they form a reasonable suspicion that the person is in possession of drugs,” the spokesman said.

“When a dog indicates that a person may be in possession of illicit drugs, this assists police in forming reasonable suspicion.”

In NSW, police may carry out “general drug detection” using sniffer dogs in certain authorised places including licensed premises, public transport, and sport and entertainment venues.

As with Victoria, “general drug detection” in NSW relates to the detection of drugs that are in the possession or control of a person.

Can the dogs tell the difference between possession and the residual scent of drugs?

Dave Wright, a former NSW Police dog trainer, told Fact Check that drug detection dogs were trained through a process of conditioning to recognise and indicate the odour of prohibited drugs.

Mr Wright, who is now the Director of Training at Dog Force Australia, a private specialist dog training company, explained that the dogs would not be able to tell the difference between a residual scent and a scent from someone in possession of drugs.

“They’re incredibly sensitive [to smell]. If someone has been carrying drugs, if they’ve had contact with drugs, even things like carrying currency that’s been sitting with drugs — the dogs will often pick up on that,” Mr Wright said.

“The sensitivity of the dog’s nose is sometimes some of the problem because they’re picking up some of the residual recent contact with the narcotic and they’re indicating the presence of that.”

More recent figures

A spokeswoman for NSW Greens MP David Shoebridge provided Fact Check with data on drug detection dogs in NSW dating back to 2009, which Mr Shoebridge collected via NSW Parliament.

That data shows that over the past decade, 63 per cent of searches following an indication found no drugs. In 2017, it was 61 per cent; for the first five months of 2018 this was the case in 63 per cent of searches.

South Australia is the only other state to make similar data publicly available — drugs were found in just 18 per cent of searches conducted after drug dog indications in the 2017-18 financial year.

Fact Check requested the equivalent data from police in all other states and territories. However, none were able to provide figures.

A different measure of accuracy?

In comments referenced in the 2006 NSW Ombudsman’s report (page 49), NSW Police said they believed drug-detection dogs were accurate 70 per cent of the time.

“This figure is obtained by adding the proportion of occasions when a drug is found to the proportion of occasions when no drug is found, yet some explanation is provided as to why the dog might have indicated the person searched,” the police said.

In an email, a spokesman for Victoria Police said 85 to 90 per cent of people indicated by drug-detection dogs were either in possession of drugs or admitted to having had recent contact with drugs.

A spokeswoman for the Western Australia Police Force said drug detection dogs had been trained to detect narcotics to a high level of accuracy.

“Any positive indications given by our detection dogs result in either a drug find or confirmed dead scent (a location where drugs have been removed recently and residual odour remains),” she wrote in an emailed response to Fact Check.

Issues with the police measure

In its 2006 review, the NSW Ombudsman took issue with NSW Police’s measure of drug dog accuracy.

The review revealed that 21 per cent of people who admitted to having been in contact with drugs claimed to have been in the presence of cannabis smoke only.

However, while NSW Police included this figure when determining their accuracy rate, they did not accept the proximity to other drug users as a possible explanation for a dog indicating the presence of drugs on someone.

Other admissions by people of previous contact with drugs were so remote — days, months and even years prior — that they could not reasonably explain an indication by a sniffer dog.

Dr Peta Malins, a lecturer in criminology and justice studies at RMIT University, told Fact Check that including admissions of previous contact with drugs in measuring the accuracy of sniffer dogs was problematic.

“My anecdotal experience is that people will say whatever they think is going to get them out of that situation as quickly as possible,” she said.

“If they think it will stop the police strip searching them or doing a more thorough search they may say they were using drugs earlier in the day.

“It’s hard to know how accurate these admissions are.”

Are the dogs “incorrect”?

Dr Malins said that while there may be no drugs found after an indication, that did not mean the dogs were not picking up on a residual drug odour.

“We can never know for sure if the dog was picking up on a drug that may have been very carefully hidden or whether they may be picking up on traces of a drug that a person may have used or had in their clothes or bag earlier.”

Whether or not the dogs were “incorrect” was a definitional question for Dr Malins.

“If police are using a drug dog indication to justify reasonable grounds to search, then you might argue that that’s not an accurate enough rate of actually finding drugs to justify being able to search someone on suspicion of possession,” she said.

“But the dogs tend to be very accurate in testing environments.”

The 2006 NSW Ombudsman report pointed out that the primary objective of using detection dogs was to find people in possession of drugs.

“Although some admissions may support the accuracy of drug detection dogs in picking up the scent of prohibited drugs, this should not be confused with the accuracy of the dogs detecting persons currently in possession of prohibited drugs, which is the purpose for their use.”

Principal researcher: Ellen McCutchan

The Victorian Greens claim that drug sniffer dogs are incorrect 75 per cent of the time. Fact Check finds that claim to be close to the mark. ]]>