Justifying Random Drug and Alcohol Testing in the Workplace
In a unanimous decision released Sept 28, 2017, the Alberta Court of Appeal (“ABCA”) upheld a judicial review decision which found that the majority of a grievance arbitration panel (“Majority Panel”) had improperly decided that Suncor’s random drug and alcohol testing policy was unenforceable.
The ABCA’s decision focused on the Majority Panel’s decision not to consider evidence of substance abuse issues in the workplace outside the limits of the bargaining unit. The ABCA found that a proper analysis should have looked look beyond the bargaining unit itself and considered substance abuse issues in the workplace as a whole.
In 2012, Suncor implemented random drug and alcohol testing for workers in safety-sensitive positions at some of its sites in the Fort McMurray area. The union (“Unifor”) grieved the alleged infringement of unionized workers’ privacy rights.
Both sides agreed that proper safety procedures were paramount to preventing workplace accidents and that such procedures included preventing drug and alcohol use at the workplace. What they disagreed on was Suncor’s use of random drug and alcohol testing as a part of that safety plan.
At arbitration, the Majority Panel ruled in favour of Unifor, holding that Suncor had not demonstrated sufficient safety concerns within the bargaining unit to justify random testing. The dissent concluded that there was overwhelming evidence of safety issues within the workplace, including substance abuse issues, and would have upheld the employer’s random testing scheme.
Suncor successfully applied for judicial review of the Majority Panel’s decision. The reviewing judge quashed the decision and sent it back for review by the arbitration panel. Unifor appealed this decision to the ABCA.
The ABCA focused its analysis on whether the Majority Panel was unreasonable in finding that it could only consider evidence that demonstrated substance abuse problems within the bargaining unit, and could ignore or disregard any evidence of substance abuse problems within the broader workplace. The ABCA concurred with the reviewing judge that the Majority Panel’s decision was unreasonable in this regard. Citing Abella J’s majority decision in Irving , the court explained that the test for whether random drug and alcohol testing is justified is defined in terms of whether there are special safety risks. Special safety risks could include a general problem of substance abuse within a workplace, not just within the bargaining unit.
Given that much of Suncor’s evidence related to the workplace as a whole, and did not distinguish between unionized employees, non-unionized employees, and contractors’ employees , most of its evidence was not considered by the Majority Panel. The Panel was of the view that because it did not have jurisdiction over those employees outside of the bargaining unit, such evidence was irrelevant. The Majority Panel further found that the inquiry was properly focused on the issues and risks facing the particular bargaining unit and that it was foreclosed from considering any evidence outside that bargaining unit.
The ABCA found that the Majority Panel’s decision was unreasonable in not considering evidence of workplace-wide substance abuse problems. The court stated at para. 44:
The majority conflated two issues, and went too far when it rejected Suncor’s policy argument. The first issue is the tribunal’s jurisdiction to decide only those matters arising under the collective agreement. The second issue is what evidence is relevant to whether there is a general problem of substance abuse in the workplace. These two questions are distinct. Simply because the tribunal lacked jurisdiction to impose or endorse drug and alcohol testing of non-unionized employees does not mean that Suncor’s overall experience with substance abuse in its Fort McMurray operations had no bearing on the question before the tribunal.
The ABCA found that the Majority Panel unreasonably concluded that these “jurisdictional” considerations required the arbitration board to “blind itself to logically relevant evidence” and to characterize Suncor’s evidence as being “unparticularized” to the bargaining unit and thus irrelevant. This misstated the Irving legal test and set the evidentiary bar too high. By unreasonably narrowing the evidence that it considered, the Majority Panel asked the wrong question, and therefore applied the wrong legal test. Applying the wrong legal test resulted in an unreasonable decision.
While the ABCA remarked that in certain workplaces there may be a good reason to distinguish between unionized and non-unionized employees in terms of evidence of substance abuse, that was not the case at Suncor given that unionized, non-unionized and contractor employees all worked side-by-side in integrated workforces at integrated jobsites. The ABCA agreed with the reviewing judge who noted that there was nothing “to suggest that alcohol and drug use within the bargaining unit differed in some meaningful way from that in the broader workforce” (para. 48).
The ABCA’s decision serves as an important clarification of the law in terms of how an employer can justify random drug and alcohol testing policies in a safety sensitive context. Evidence of overall workplace drug and alcohol issues will likely be relevant to the justification analysis. An employer’s evidence likely does not need to be limited to demonstrating such issues within the specific bargaining unit or employee pool to which the policy is addressed. This decision confirms that an employer has some degree of flexibility to cast a wider evidentiary net to prove its case that imposing a random drug and alcohol policy is justified.
 Of its approximately 10,000 employees working on site, 3,383 of those employees were represented by Unifor, 2,963 were non-union employees, and the remaining workers were employed by contractors.