Reprinted by permission of LexisNexis Canada Inc., from the Class Action Defence Quarterly Edited by Eliot N. Kolers, Copyright 2020.
On June 26, 2020, the Supreme Court of Canada (SCC) released its much-anticipated decision in Uber Technologies Inc v Heller.1
The appeal considered whether David Heller’s class action, commenced against Uber in 2017 on behalf of its taxi and food delivery drivers for violations of Ontario’s Employment Standards Act (ESA),2 ought to be stayed before the Ontario court and referred to arbitration pursuant to a mandatory arbitration clause contained in Uber’s standard-form driver services agreement.
In an 8-1 decision, the SCC held that in cases where a party’s contractual rights are “illusory”3 due to inequality of bargaining power or unduly onerous terms which render resolution of a matter by arbitration “realistically unattainable”,4 that party’s right to seek a resolution through the courts—including by way of class action litigation—will be upheld.
The governing law in Canada is that an arbitrator should be the first to decide on preliminary challenges to an arbitrator’s jurisdiction where an agreement to arbitrate is arguably in place. The SCC held, however, that in cases where there is a genuine challenge to arbitral jurisdiction and a “real prospect” that referral to arbitration may result in a challenge never being resolved,5 access to dispute resolution through court proceedings must remain available to litigants.
The flood of publicity arising from this appeal speaks to its perceived influence on a broad spectrum of legal matters, including class actions, international arbitration, business contracts, and employment law. The decision’s impact on the use of standard form contracts in the growing international gig economy is of particular concern for multinational companies and big tech, where reliance on mandatory arbitration provisions to manage litigation and class action exposure across borders has become customary.
While concurring Justice Brown6 and dissenting Justice Coté,7 along with numerous commentators, expressed concerns with the majority’s ruling, Uber brings some firm and important conclusions to the Court’s ongoing efforts to resolve clashes between arbitration provisions, contractual principles, legislative protections for categories of litigants, and class proceedings in Canada. As explored in my previous article for this newsletter in June 2019,8 cases such as ZI Pompey Industries v ECU-Line NV, [2003] S.C.J. No. 23, Douez v Facebook, Inc, [2017] S.C.J. No. 33, and TELUS Communications Inc v Wellman, [2019] S.C.J. No. 19, display the Court’s historical efforts to reconcile these intersecting regimes in an increasingly globalized business climate.
On the basis of this history, speculations that Uber could uproot the arbitration framework and cause contractual uncertainty are inflated. By and large, the effect of the SCC’s decision in Uber is to establish clearer boundaries between court and arbitral jurisdiction on stay motions in Canada.
As I detail below, the critical takeaway from Uber is that, in order to be effective, arbitration agreements must be drafted to provide realistic and effective access to recourse to both parties, particularly in the context of standard form contracts. This outcome flows logically from the fundamental public policy concerns of efficiency, predictability, and access to justice that collectively underlie the laws of arbitration, class actions, contract, and civil procedure. Uber situates these policy objectives within the gig economy and broader contemporary business environment.
Canadian courts have introduced increasingly strong protections for arbitration into the law of contracts, based on it being a cost-effective and efficient method of resolving disputes. Both Ontario’s Arbitration Act, 1991 and International Commercial Arbitration Act, 2017 (ICAA) mandate a stay of court proceedings where a dispute is subject to an agreement to arbitrate.9 While previous decisions interpreting the mandatory stay provisions such as Pompey,10 Facebook,11 and TELUS12 had distinct features, all three upheld the general rule that forum selection and arbitration clauses are presumptively enforceable, based on the policy of promoting efficiency, predictability, and access to justice through resort to alternative dispute resolution.
Further to the basic mandatory stay provisions, both the domestic Act and the ICAA adopt the international principle of compétence-compétence, which provides that an arbitral tribunal is competent to determine is own jurisdiction.13 This principle instructs that even preliminary objections to arbitration agreements should be stayed before the courts and referred to the tribunal.
The Ontario arbitration statutes implement compétence-compétence through the combined operation of: (a) the express statutory jurisdiction of the tribunal to rule on its own jurisdiction—specifically, to rule on objections about the “existence or validity of the arbitration agreement”;14 and (b) the supporting provisions that require court proceedings to be stayed if the parties have agreed to submit the dispute to arbitration, with specified exceptions (including for an invalid arbitration agreement).15
Limited exceptions to the automatic rule of referral are carved out for both the mandatory stay provisions and compétence-compétence in the Ontario legislation. Under the ICAA, a court may refuse to refer a dispute to arbitration if the arbitration agreement is “null and void, inoperative or incapable of being performed”.16 Section 7(2) 2 of the domestic Act adds that the court may refuse to grant a stay of court proceedings if “the arbitration agreement is invalid”.17 As a result of these exceptions, courts in Ontario are provided a window of concurrent jurisdiction to decide on preliminary objections to arbitration agreements under certain conditions.
This overlapping jurisdiction presented clear challenges for courts in responding to stay motions. In Union des consommateurs c. Dell Computer Corp (Dell), the SCC introduced a “general rule” of “systemic referral to arbitration” to assist courts in exercising this concurrent authority. Here, the SCC established that courts should depart from the general rule only if the challenge is:
Further, Dell provided the court must be satisfied that the challenge is not a delay tactic and it will not unduly impair the arbitration proceeding prior to departing from the rule of general referral.19
Following Dell, the law preceding Uber was that arbitral tribunals lacked exclusive authority to determine challenges to jurisdiction at first instance in Ontario—courts retained concurrent authority in limited situations.20
The plaintiff, David Heller, commenced a class proceeding against Uber in 2017 in Ontario for violations of the ESA. At that time, Mr. Heller worked as a food services and delivery driver in Toronto using Uber’s software applications.
To become a driver for Uber, Mr. Heller was required to accept the terms of Uber’s standard form services agreement. Under the terms of that agreement, drivers were required to resolve any disputes with Uber through mandatory arbitration in the Netherlands.21 The agreement stated that arbitration would be conducted in accordance with the ICC rules, which require up-front administrative and filing fees of USD$14,500, plus all other legal fees and costs of participation.22 The undisputed evidence regarding Mr. Heller’s income as an Uber driver was that he earned approximately $400-$600 per week based on 40 to 50 hours of work, or $20,800-$31,200 per year, before taxes and expenses.23 Accordingly, the costs to arbitrate a claim against Uber would amount to all or most of the gross annual income he earned working full-time as an Uber driver.
Uber brought a motion to stay the class proceeding in favour of arbitration in the Netherlands, relying on the arbitration clause.24 Justice Perell of the Superior Court stayed the proceeding, holding that the arbitration agreement’s validity had to be referred to arbitration, in accordance with s. 17(1) of the domestic Act and Article 16(1) of the ICAA.
The Court of Appeal unanimously reversed the stay, finding that the agreement was invalid because it: (a) constituted an impermissible contracting out of the ESA’s protections; and (b) was unconscionable, based on the inequality of bargaining power between the parties, and the “improvident” (or unreasonable) cost of arbitration.25
Uber presented two novel challenges for the SCC in applying the existing arbitration law in Ontario:
Pursuant to Dell, if a challenge to the validity of an arbitration agreement raises a question of law or mixed law and fact requiring only a superficial review of the evidence, the court may decide on the contractual issues presented on a motion for a stay.
A ruling on this basis alone would have fit squarely within the institutional role and core competency of the Ontario courts established by the legislature, particularly when the interpretation and effect of an Ontario statute (the ESA, in this case) is at issue. In a proposed class proceeding involving an arbitration clause in a contract of adhesion, a court’s ruling may also provide a common and binding (or highly persuasive) resolution of a legal issue applicable to all parties to the arbitration clause, promote access to justice, and ensure a consistent resolution of the legal question—whether the matter ultimately remains in court or is referred to arbitration.
In the face of an arbitration agreement that presented glaring obstacles to one party’s access to justice, the SCC in Uber was tasked with applying the rule in Dell while maintaining the core policy concerns of efficiency, predictability, and access to justice that underlie Ontario’s arbitration framework. In fact, these objectives are also essential to the ESA,26 the Class Proceedings Act,27 and the civil courts’ approach to public policy generally.28
As stated by the majority, “Dell did not contemplate a scenario wherein a matter would never be resolved if the stay were granted. Such a situation raises obvious practical problems of access to justice that the Ontario legislature could not have intended when giving courts the power to refuse a stay.”29
The absence of bona fide access to recourse became the majority’s core concern in Uber. The Court held that “when arbitration is realistically unattainable, it amounts to no dispute resolution mechanism at all”.30 Because Mr. Heller would not possibly have the resources to arbitrate his issues under Uber’s requirements (namely, the costs and travel associated with arbitration in the Netherlands), he and other drivers would have no reasonable option of having their complaints resolved.
This position closely mirrors concerns identified by Justices Abella and Karakatsanis in their dissenting opinion in TELUS (2019).31 There, consumers in the prospective class were automatically exempt from a mandatory arbitration clause in TELUS’s mobile services agreements pursuant to Ontario’s Consumer Protection Act.32 However, in their opinion, the exclusion of TELUS’s business customers from the class risked preventing them from pursuing their claims for overpaid fees entirely, precluding access to justice for low-value claims that could not justify the expense of individualized arbitration33 and “further undermin[ing] Ontario’s class actions regime as a viable, procedural access to justice mechanism”.34
Focusing on the ultimate objective of ensuring true access to recourse, the majority in Uber found the arbitration agreement to be unenforceable on two bases.
First, the Court held that it had jurisdiction to decide on the challenge to the arbitration agreement because the facts of the case raised an issue of “accessibility” that was not presented in Dell and justified departing from the systemic rule of referral. While the majority found that the conditions set out in Dell could indeed be satisfied to justify a determination by the Court on the invalidity of the agreement, the court went further to establish a new ground for court jurisdiction to refuse a stay where practical access to justice concerns are presented. Specifically, this exception would apply in cases where access to a decision by an arbitrator was unlikely or impossible.
The Court set out that: “To determine whether only a court can resolve the challenge to arbitral jurisdiction, the court must first determine whether, assuming the facts pleaded to be true, there is a genuine challenge to arbitral jurisdiction. Second, the court must determine from the supporting evidence whether there is a real prospect that, if the stay is granted, the challenge may never be resolved by the arbitrator”.35
Next, having established this jurisdiction, the Court found Uber’s arbitration agreement to be invalid because it was unconscionable. The majority again followed direction from the TELUS case that “arguments over any potential unfairness resulting from the enforcement of arbitration clauses contained in standard form contracts are better dealt with directly through the equitable doctrine of unconscionability”,36 which it said allows the court to “invalidate a single clause within an otherwise enforceable contract”37 in order to resolve grossly unfair or unreasonable terms.
Given inconsistencies in the application of the doctrine by lower courts, the majority sought to explore and clarify the approach to unconscionability, which it held to be a two-step test, composed of: (1) proof of inequality in the positions of the parties; and (2) proof of an improvident bargain.38 This was consistent with Justice Abella’s concurring judgment in Facebook, where she applied the two-step approach to a forum selection clause in a standard form consumer contract that she said created an “unfair and overwhelming” benefit for the drafting party, Facebook.39
Unconscionability is perhaps the most disputed component of the majority’s reasons. Justice Brown, in his concurring opinion, would have found the agreement invalid as a “simple” matter of public policy that “courts will not enforce contractual terms that, expressly or by their effect, deny access to independent dispute resolution”.40 A provision “that penalizes or prohibits one party from enforcing the terms of their agreement”, he states, “directly undermines the administration of justice”.41 In Brown J.’s view, the application of the unconscionability doctrine was “entirely unnecessary”, and resulted in what he felt was a “drastic expansion” of the doctrine’s scope.
In her dissenting opinion, Justice Côté disagreed with the application of both the unconscionability and public policy doctrines to Uber’s arbitration clause, noting:
Although times change and conventional models of work and business organization change with them, the fundamental conditions for individual liberty in a free and open society do not. Party autonomy and freedom of contract are the philosophical cornerstones of modern arbitration legislation.42
Applying Dell, she found that the doctrine of unconscionability and the ESA raised questions of mixed law and fact that could not be decided on a superficial review of the evidence and should therefore have been referred to an arbitrator. In her typical fact-driven approach, Justice Côté found that the terms of the arbitration agreement were made clear to Mr. Heller at the time of signing the contract, and the agreement, having been freely entered into, ought to be upheld. Notably, however, she still found the arbitration fees to be untenable for Mr. Heller. She therefore would have granted a stay of proceedings, on the condition that Uber advanced the funds needed to initiate the arbitration.
Multinational organizations conducting business in the globalized marketplace rely on standard form agreements to structure their activities effectively and efficiently. Understandably, such organizations also seek to draft agreements which limit and foster predictability in their litigation exposure, particularly in terms of class actions. However, this process presents a clear risk where a party with significantly greater resources and know-how seeks to restrict or eliminate the path of dispute resolution evolving from those contracts. Should an issue arise under that agreement, the prescribed recourse may not be genuinely understood by contracting individuals at the time of formation, and may not be viable, practical, or attainable in the event a dispute arises.
The majority’s decision in Uber makes clear that the availability of access to meaningful legal recourse will be paramount to the courts on stay motions. Individuals contracting with businesses in the gig economy will be guaranteed access to dispute resolution through the courts if the circumstances of their agreements make arbitration impossible:
Access to civil justice is a precondition not only to a functioning democracy but also to a vibrant economy, in part because access to justice allows contracting parties to enforce their agreements. A contract that denies one party the right to enforce its terms undermines both the rule of law and commercial certainty.43
Uber does not render arbitration agreements worthless in the context of standard form contracts. Rather, companies contracting with workers and other service providers in the gig economy, or deploying mass contracts of adhesion generally, should be cognizant when drafting mediation and arbitration provisions that those agreements provide a realistic and effective opportunity for dispute resolution. As the majority states, the purpose of arbitration agreements is not to insulate a party from any meaningful challenge, nor to deny the other any form of relief.44
Following Uber, companies who fail to consider access to justice and relevant statutory dispute resolution mechanisms when drafting their standard form agreements may risk exposure to individual as well as class action lawsuits, and possibly, a determination by the courts that the agreements are invalid on the basis of unconscionability.
In the wake of the SCC’s decision, judges, practitioners, and businesses can expect to see a rise in litigation concerning “accessibility” issues and arbitration agreements, as well as unconscionability claims where standard form agreements and other contracts of adhesion are involved.
Heller will now be allowed to proceed with a $400 million class action on behalf of Uber drivers across Canada. The parallel U.S. class action, O’Connor v Uber, was settled for $20 million in March 2020.45
Notes