Newfoundland and Labrador – Labour Arbitrator’s Collective Agreement Interpretation Passes Vavilov Reasonableness Muster – #559

In Pennecon Maintenance Services Limited v. Fish, Food & Allied Workers, 2021 NLSC 141, Justice Knickle ruled that a labour arbitrator reasonably interpreted a collective agreement in light of the precepts laid down in both Canada (Minister of Citizenship and Immigration) v. Vavilov, 2019 SCC 65 [Vavilov] and Sattva Capital Corp. v. Creston Moly Corp., 2014 SCC 53 [Sattva]. Although arising from an application for judicial review, Justice Knickle’s analysis provides relevant insights applicable to private arbitration disputes as they relate to contract interpretation.

Vale Newfoundland and Labrador Limited (“Vale”) retained the Applicant, Pennecon Maintenance Services Limited (“Pennecon”), to perform maintenance and repair work at the Long Harbour Processing Plant in Long Harbour, Newfoundland and Labrador. Pennecon’s employees were members of the Fish Food and Allied Workers union (“FFAW”).

Pennecon and FFAW entered into a collective agreement (“CA”). One of the CA’s terms required certain Pennecon employees (Maintenance Support Technicians) to “provide training and knowledge transfer” to Vale’s employees. These Pennecon employees’ training work entitled them to an hourly bonus on top of their regular wages.

Vale advised Pennecon in Spring 2019 that it no longer required these training and mentoring services and would stop paying the premium, which Pennecon subsequently paid to its workers under the CA. Pennecon in turn advised it would stop paying its employees this additional compensation. FFWA grieved this decision.

According to Justice Knickle, the key issue in the arbitration was the proper meaning of the “mentoring uplift clause”, the clause in the CA addressing the bonus payable to Pennecon’s employees for training and mentoring. That clause read as follows:

Maintenance Support Technicians have a responsibility to mentor, coach and train Vale Processing Plant Operators in maintenance activities, as well as each other, to create a multi-skilled non-jurisdictional learning environment. In recognition of this expectation, the Company will pay a mentoring uplift of four dollars fifty five cents ($4.55) per hour for each regularly scheduled hour worked. The Mentoring Uplift will not be paid to Maintenance Repair Technicians

The Parties agreed before the arbitrator that he had to consider not only the CA’s language, but also the circumstances surrounding its formation. They disagreed, however, on the import of the words “each other” in the uplift clause.

Pennecon argued the clause applied only to the specific mentoring services its employees provided to the Vale employees, which had ceased, not to mentoring and training as between its own employees. For its part, FFWA advocated a broader meaning that covered mentoring and training activities as between Pennecon’s own employees. Under this interpretation Vale’s decision to discontinue Pennecon’s mentoring of the Vale employees did not exhaust the mentoring uplift clause’s application. FFWA based this argument on evidence that the mentoring activities in fact occurred not only between Pennecon and Vale employees, but also among Pennecon employees.

In reaching his conclusion, the arbitrator found he had to assign meaning to the words “each other”, and that Pennecon’s interpretation failed to do so. He noted the “stated goal” of the CA’s relevant provisions was “the creation of a multi skilled, non jurisdictional learning environment”. Extending the mentoring uplift clause to interactions between Pennecon employees was, he found, consistent with that goal. In the result, the arbitrator concluded Pennecon could not discontinue the mentoring uplift payments without first advising its employees that they were no longer expected to provide mentoring and training to “each other”.

On judicial review, Justice Knickle found no reason to rebut the presumptive reasonableness standard of review. Indeed, the Parties agreed on this point. She then summarized reasonableness review’s characteristics as the majority Justices of the Supreme Court of Canada in Vavilov explained them:

[22] The majority in Vavilov explained that the review of a decision on the standard of reasonableness is not a question of what this Court would have done if tasked with making the decision, but whether “the decision made by the administrative decision maker – including the rationale for the decision and the outcome to which it led – was reasonable” (Vavilov at paragraph 83).  Both the process of the decision maker or the process in the making of the decision, as well as the justification for the decision, are subject to the reasonableness assessment.

[23] The reasonable decision is characterized by three essential qualities: justification, transparency and intelligibility (Vavilov at paragraph 99).  The majority for the Court in Vavilov stated the review of the decision is not a “line by line treasure hunt” for error (paragraphs 102-104), but the review court must be satisfied there are no “fatal flaws”  in the “overarching logic”.

Justice Knickle went on to note Vavilov’s focus on the need for a decision to exhibit internal coherence within the legal and factual constraints in play.

After describing the reasonableness analysis, Justice Knickle recognized that since the matter before the arbitrator was one of contractual interpretation, the “legal constraints” fencing in the arbitrator’s exercise start with the Supreme Court of Canada’s earlier decision in Sattva:

  1. Contract interpretation requires the decision-maker to read the contract’s terms in light of the factual matrix;
  2. The factual matrix consists of the “objective evidence of the background facts at the time of the execution of the contract” that both parties knew or reasonably ought to have known; and
  3. The factual matrix can never overwhelm the contract’s terms.

Applying these principles, Justice Knickle found no basis to interfere with the arbitrator’s interpretation. She observed no fatal flaw in the arbitrator’s chain of reasoning. She also found the award sufficiently transparent and intelligible, and that it adequately justified the outcome.

Justice Knickle also noted the arbitrator’s “sensitivity to the labour context was evident throughout his decision” and how he considered that Pennecon and FFWA (and its members) were in a continuing long-term relationship. To that end, the arbitrator did not leave Pennecon without options; he specifically stated Pennecon would no longer have to make the mentoring uplift payments upon advising its workers they were no longer expected to mentor and train each other.

Finally, Justice Knickle rejected Pennecon’s argument that the arbitrator failed to consider its witness’s evidence. She found the arbitrator considered and rejected that evidence, which would have deprived the “each other” language in the mentoring uplift clause of meaning. This, she said, would have at best caused the context to overwhelm the CA’s text, an error of law.

Contributor’s Note

First, and as alluded to above, this case provides some useful guidance applicable to private arbitration disputes even though it arose in the administrative law context. Labour arbitrator decisions have long been treated under public law. However, in interest arbitrations (i.e., disputes regarding a collective agreement’s terms and conditions) such as this, the arbitrator’s task is to interpret a contract. The same body of law, including Sattva, applies to all contract interpretation, regardless of the surrounding private law or public law setting. For another example of an administrative law decision informing the private dispute-resolution context, see the note in Federal CA – Arbitrator/Adjudicator expressing “tentative views” in pre-adjudication mediation to foster settlement not indicative of bias – #515.

Second, in reviewing the excerpts of the arbitrator’s award contained in the decision, it seems he might have assimilated the parties subsequent conduct—facts relating to the contract’s performance—within his assessment of the factual matrix—facts relating to and occurring at or before the contract’s formation. The Court of Appeal for Ontario pointed out the important distinction between these two concepts in Shewchuk v. Blackmont Capital Inc., 2016 ONCA 912. In that case, the Court of Appeal explains that while the factual matrix is, per Sattva, a necessary part of the contract interpretation exercise, considering subsequent conduct is not. The Court also warned this evidence can in some cases be inherently unreliable, specifically where a party relies on its own self-serving conduct. This and other considerations caused the Court of Appeal to conclude that “[e]vidence of subsequent conduct should be admitted only if the contract remains ambiguous after considering its text and its factual matrix”.

From the decision in Pennecon, it is unclear whether the arbitrator found the mentoring uplift clause ambiguous, allowing him to consider subsequent conduct. If not, this would arguably have been a legal error or, in Vavilov parlance, a failure to decide within the legal constraints of contract interpretation law. It is unclear from Justice Knickle’s reasons whether this was raised, either before her or before the arbitrator below.