Drug Detection Dogs Fail Smell Test

May 2015

Article by: Kim Thorne

Previously printed in the CCH’s Focus on Canadian Employment and Equality Rights (CEER) Newsletter.

In United Steelworkers Local 7552 v. Agrium Vanscoy Potash Operations (Grievance 16-10, Random Drug Searches/Interviews), [2015] S.L.A.A. No. 1 (Norman), a Saskatchewan employer’s practice of periodically deploying drug detection dogs to screen individuals entering its safety-sensitive worksite was held to be in breach of the collective agreement.

Impugned process

The employer, Agrium Vanscoy Potash Operations, utilized a security service with drug sniffing dogs to screen persons entering its mine site. This occurred at the site guardhouse on randomly selected days. This mandatory entry requirement included follow-up interviews of individuals who triggered an “alert” from one of the dogs. A second sniff test of the subject and his or her possessions would then take place. If illegal substances were found or the individual gave other reasonable cause, a manager could advise that a urinalysis test be conducted at an offsite lab for the detection of drug metabolites in the individual’s system. The union grieved this “search/interview” process as, in its view, the process amounted to an unjustifiable violation of the fundamental employee right to privacy.


  • Arbitrator Ken Norman accepted that the mine is a safety-sensitive workplace and began his analysis by drawing a distinction between drug testing cases where bodily substances are seized and tested and the present case where a “search/interview” process was at issue. He also noted that unilateral random deployment of drug detection dogs to screen employees entering a safety-sensitive workplace had never previously been considered under the reasonableness test in K.V.P. Co. (1965), 16 L.A.C. 73, a test which was central to the Supreme Court of Canada’s analysis in Communications, Energy and Paperworkers Union of Canada, Local 30 v. Irving Pulp & Paper, Ltd., [2013] 2 S.C.R. 458.

Charter values

Arbitrator Norman held that the question of whether a unilateral company rule which intrudes on employees’ privacy is reasonable must be examined through the lens of “Charter values”. In so doing, the arbitrator imported into the analysis section 8 Charter jurisprudence out of the Supreme Court of Canada. This jurisprudence considers the underlying values of dignity, integrity and autonomy. Because employees who triggered an alert from a dog were required to explain themselves in a backroom, the arbitrator concluded that this constituted a serious intrusion upon their reasonable expectation of informational privacy or their “biographical core”. A forced explanation could, in the arbitrator’s view, reveal intimate information about the employee’s lifestyle.

Threshold question and balancing of interests

Arbitrator Norman referred to Irving’s balancing of interests approach in weighing the privacy concerns of the employees against the safety concerns of the employer. The arbitrator held, however, that before such a balancing approach could be applied, there had to be an initial threshold consideration of whether there was an actual safety problem justifying the policy:

The prior threshold stage in the justificatory argument limiting rights under the Charter sets the bar very high; calling for proof of a pressing and substantial objective demonstrably justifiable in a free and democratic society, for the challenged measure. Under “Charter values” analysis, I take the threshold bar to have been set by Irving as “evidence of enhanced safety risks, such as evidence of a general problem with substance abuse in the workplace” ….

[Emphasis added.]


In this case, the employer did not adduce sufficient evidence to establish a general problem of drug use at its worksite and thus fell short of meeting what Arbitrator Norman saw as the threshold standard set in Irving. The arbitrator ruled that factors which do not form part of the threshold analysis such as deterrence, detection and risk reduction did not come into play, and went on the conclude that the employer’s “search/interview” process was in violation of the collective agreement.

Interestingly, it was not the drug sniffing dogs per se that Arbitrator Norman found to be objectionable. Even though the balancing of interests analysis was not performed by the arbitrator, he opined that the minimal intrusiveness of the brief sniffing by a dog was to be weighed in the balance together with the more intrusive lifestyle-related questioning which would ensue in the backroom. The dogs were the means by which that objectionable component was triggered.

With all due respect, Arbitrator Norman’s reasoning in this decision is flawed in at least a couple of respects. It appears to represent a wholesale importation of Charter analysis into private contract law. At some point, there will no doubt be judicial consideration of this.

As well, the arbitrator’s decision is silent with respect to the many arbitration decisions over the last decade or so which stand for the proposition that in safety-sensitive workplaces, the employer need not advance proof of an actual drug problem before adopting drug testing policies (which are of course far more intrusive than the use of drug sniffing dogs). The reality is that employees who use drugs or alcohol at work do so illicitly and covertly and try very hard to avoid detection, making proof of an actual problem notoriously difficult to prove by any means. Judicial consideration of the threshold test articulated by Arbitrator Norman will be welcome.


This case appears to be part of a growing trend in the case law – the requirement for proof of an actual problem at the worksite. Employers who are considering the use of drug sniffing dogs to detect the movement of drugs onto a worksite should carefully do their homework and compile as much evidence of drug use at work as possible. Consider working with the EFAP provider and trying to obtain data which does not reveal the identity of employees but speaks to the frequency of drug or alcohol treatments among employees at the site. Document all incidents of drug or alcohol use including the smell of marijuana and any discovery of drug paraphernalia, and think about using anonymous surveys regarding drug use at work.

Employers may also want to engage a specialist to conduct on-site detection of drug residue through carefully administered tests on work stations, equipment, controls, etc.

Adoption of measures such as drug detection dogs will likely be found unreasonable in the absence of proof of an actual drug issue at the workplace.

Employers should also refrain from insisting that employees who are suspected of bringing drugs onto the worksite be subjected to drug testing. The drug test will not help prove that the employee was impaired at work. Keep policies about impairment separate from policies aimed at preventing movement of drugs into the workplace.

It is not clear what the appropriate standard should be for introducing drug sniffing dogs into a “search/interview” process, but the analysis should at a minimum recognize the distinction between random tests for impairment and random searches for possession.

We would expect that this decision will be judicially reviewed. Until then, before implementing a similar policy, employers should critically consider whether there is actually evidence of enhanced safety risks due to drug use in their workplace.