At the end of June, the Supreme Court of Canada released one of its most important contract law decisions of the last decade: Uber Technologies Inc. v. Heller [“Uber”]. The issue in Uber was whether the arbitration clause in the standard Uber services agreement was valid. The clause required the arbitration of any dispute through mediation and arbitration in the Netherlands, along with an up-front administrative and filing fee of US $14,500. There was no dispute that filing fee represented the majority of the plaintiff’s annual income.
Eight of the nine justices of the Supreme Court voted to invalidate the arbitration clause, with only Justice Côté voting to uphold it. Of the eight, seven took the position that the clause was “unconscionable” while one – Justice Brown – argued against relying on the doctrine of unconscionability and took the view that the clause was contrary to public policy. In this post, I argue that Brown J.’s concurrence is correct as a matter of law. It reaches an intuitively satisfying conclusion and does so by appealing to settled common law doctrine. The majority’s decision, by contrast, is likely to bring uncertainty and instability to contract law doctrine.
The Majority’s Decision in Uber
The plaintiff brought a class action against Uber for violating employment standards legislation. Uber’s standard form employment contract required all disputes to be mediated and arbitrated in the Netherlands. There was no dispute that the process would require an upfront administrative and filing fee of US $14,500. There was also no dispute that the plaintiff earned anywhere from $20,800 to $31,200 per year before taxes, and that the cost to arbitrate a claim against Uber equaled all or most of his gross annual income.
The parties disagreed as to whether the case was governed by the Arbitration Act or the International Commercial Arbitration Act [“International Act.“] The latter applies to agreements that are “commercial” in nature. Both the majority and the concurrence agreed that an employment agreement is not the sort of “commercial” agreement that the International Act was intended to capture with Justice Côté taking the view that the International Act applied. The parties also disagreed as to whether the court was empowered to decide whether the arbitration agreement was “invalid” pursuant to s.7(2) of the Arbitration Act, or whether this issue had to be resolved by the arbitrator in the Netherlands. Once again, the majority and Justice Brown took the former view while Justice Côté took the latter.
These preliminary jurisdictional issues are fascinating in their own right, but in this post I intend to focus solely on the substantive issue: whether the arbitration agreement was in fact invalid. The majority found that it was and relied on the doctrine of unconscionability.
Unconscionability is an equitable doctrine that has arises where there is an inequality of bargaining power. Like other equitable doctrines, however, determining whether a bargain is unconscionable is not a free-standing results-driven inquiry. It has traditionally been applicable only in very particular circumstances and always in a principled fashion. The classic example of the doctrine of unconscionability is the ship captain stranded at sea who agrees to pay an extortionate price for his rescue.
The majority took the view that unconscionability requires both an inequality of bargaining power and a resulting improvident bargain. Crucially, however, the majority held that the doctrine does not require an overwhelming inequality, nor does it require knowledge on the part of the more powerful party as to the vulnerability of the weaker party. Some degree of inequality is sufficient provided that the resulting bargain is also improvident. On the facts of this case, there was clearly an inequality of bargaining power between Uber and the plaintiff since the plaintiff was not able to negotiate his employment contract. The bargain was also clearly improvident since the arbitration clause imposed a US $14,500 barrier, which was close to the plaintiff’s annual income and grossly disproportionate to the amount at issue in the arbitration.
Justice Brown’s Concurrence
Justice Brown staunchly disagreed with the majority’s characterization of the doctrine of unconscionability. Whereas the majority portrayed it as a doctrine resting on twin pillars – inequality of bargaining power and improvident bargain – Justice Brown noted that the case-law and commentary focus far more heavily on the former. Unconscionability is by and large a procedural doctrine that is concerned with how the contract was formed. The improvidence of the bargain itself is certainly relevant, but mostly because it points to a procedurally flawed transaction. In other words, the issue is not whether the bargain is improvident per se, but whether the improvidence of the bargain evidences an inequality of bargaining power.
Since the doctrine is mostly procedural in nature, there must be more than a mere inequality of bargaining power. Indeed, there must be a particular vulnerability on the part of the plaintiff, such that she can no longer take care of her own self-interest. The more powerful party must also have knowledge of the weaker party’s vulnerability. There need not be malice or an intent to injure, but there must be at least constructive knowledge of the vulnerability.
On the facts, there was no evidence of any particular vulnerability on the part of the plaintiff. The implication of the majority’s position is that all of Uber’s employment contracts are unconscionable, regardless of the particular strengths or vulnerabilities of the employees who sign them, merely because the terms cannot be negotiated. The Supreme Court had never previously endorsed the notion that a standard form contract could be procedurally unconscionable, but going forward all standard form contracts will be subject to be reviewed on substantive reasonableness grounds by courts that are ill-suited for the task.
There was also no evidence that Uber possessed knowledge of any vulnerability on the part of the plaintiff. And how could there have been? The plaintiff and Uber entered into a standard form contract and had no dealings beyond that. Brown J. noted, correctly in my view, that by eliminating the knowledge requirement, the court was also undermining the principles of finality and predictability. Contracting parties are “left to wonder whether an unknown state of vulnerability will someday open up their agreement to review on grounds of ‘fairness’.”
For Brown J., however, finding that the agreement was not unconscionable did not end the analysis. Despite the fact that the doctrine of unconscionability did not apply, the arbitration agreement was still void for “public policy”. Justice Brown discussed the history of this doctrine, which the Supreme Court famously discussed in Re Millar Estate. In that case, Chief Justice Duff explained that a contract could only be voided on public policy grounds where the harm to the public was “substantially incontestable” and did not depend upon the “idiosyncratic inferences of a few judicial minds.” Public policy has long been thought to be a form of “illegality” and thus it arguably goes to the very validity of the contract. Where parties purport to oust the supervisory jurisdiction of courts or another neutral arbitrator, the courts have typically found the contract to be injurious to the justice system and, more broadly, to the rule of law, and have thus invalidated the agreement on public policy grounds (for a further discussion of the doctrine, see this excellent article written by Brandon Kain and Douglas Yoshida).
If, for example, an employment contract states that the employee has no legal recourse if the employer breaches the contract, this contract would be invalidated on the grounds that it is contrary to public policy. Even if the employee were sophisticated and understood the entirety of the agreement, this particular term would be invalidated. The rule of law and the proper administration of justice require that parties are able to litigate their disputes and vindicate their rights. Put another way, contracting parties cannot agree to do away with the very system that provides for the enforcement of contracts in the first place.
In this particular case, while the arbitration agreement did not expressly oust judicial or arbitral review, it did so in effect by making the arbitration practically inaccessible. It was not simply that the plaintiff was being discouraged from advancing low value claims. Rather, “[t]he costs of proceeding to arbitration are so prohibitive that the agreement effectively bars any claim that Mr. Heller might have against Uber” (para. 132). The arbitration agreement, in other words, was in fact a non-arbitration agreement, which was contrary to the rule of law and, thus, to public policy. “It is,” as Brown J. quipped, “the rule of law, not the rule of Uber.”
Justice Côté’s Dissent
Justice Côté agreed with Brown J.’s analysis on the doctrine of unconscionability, but disagreed on the application of the public policy doctrine. In her view, Brown J.’s application of the doctrine represents an old-fashioned hostility to arbitral justice, which has been displaced by a more modern approach in which arbitration is encouraged and accepted as an alternative form of dispute resolution. For Côté, “[p]arty autonomy and freedom of contract are the philosophical cornerstones of modern arbitration legislation” (para. 177) , and this militates in favour of a generous approach that facilitates rather than frustrates the arbitration process.
Thus, rather than invalidating the arbitration agreement, courts can order a conditional stay or sever the illegal provisions from the contract. Côté J. would therefore have ordered Uber to advance the filing fees of the arbitration as a condition of the stay. Alternatively, Côté J. would have re-written the arbitration agreement as follows (see para 334):
Any dispute, conflict or controversy, howsoever arising out of or broadly in connection with or relating to this Agreement, including those relating to its validity, its construction or its enforceability, shall be first mandatorily submitted to mediation proceedings
under the International Chamber of Commerce Mediation Rules (“ICC Mediation Rules”). If such dispute has not been settled within sixty (60) days after a Request for Mediation has been submitted under such Mediation Rules, such dispute can be referred to and shall be exclusively and finally resolved by arbitration under the Rules of Arbitration of the International Chamber of Commerce (“ICC Arbitration Rules”). The ICC Rules’ Emergency Arbitrator provisions are excluded. The dispute shall be resolved by one (1) arbitrator to be appointed in accordance with the ICC Rules. The place of arbitration shall be Amsterdam, The Netherlands.
As Brown J. noted, however, the illegality of the arbitration agreement arose not from any particular provision but from the cumulative effect of the agreement. What is more, severing any one provision would fundamentally alter the bargain. Similarly, granting a conditional stay would force Uber into a position it clearly would have never intended – being compelled to front all arbitration costs regardless of how frivolous the claim. Thus, rather than giving effect to the principles of party autonomy and freedom of contract, Côté J.’s approach would actually harm these principles by rewriting the bargain in a manner that neither party would have intended.
Analysis of the Decision
As should be apparent, I believe Justice Brown got it right. The doctrine of unconscionability is not a license for judges to pronounce on the wisdom or fairness of a contract. It is a fail safe that permits courts to override a bargain in narrow circumstances where one of the parties is unable to safeguard her own interests. Justice Brown’s conception of unconscionability as a procedural doctrine would foster the values of predictability and consistency, as lawyers would be able to properly advise their clients at the time of contract formation as to whether the contract is likely to hold up in court. By contrast, the majority’s more substantive and expansive articulation of the doctrine will exacerbate legal indeterminacy. Wherever a power imbalance of some kind is present, lawyers will have to advise their clients of the risk, however small, that a court will invalidate the agreement. The majority does not propose a definitive framework for when the doctrine will be applied and so ultimately this will be left, in large measure, to the discretion of individual motion and trial judges.
In sum, the majority has transformed the doctrine of unconscionability from one that concerned the exploitation of vulnerability to one that pronounces on the ultimate fairness of particular contractual provisions.
And as Brown J.’s concurrence discusses at length, there was no need for the Court to go down this road. The well-established doctrine of public policy already mandates that parties cannot place their agreements beyond the review of a court or other neutral arbiter. Unlike the doctrine of unconscionability, which in its expanded form is highly discretionary, the doctrine of public policy is narrowly confined to certain core principles such as the rule of law and the administration of justice. Côté J. is certainly correct that the court’s historic hostility to arbitration has evolved and has been supplanted by legislation that expressly permits and facilitates arbitration. However, the courts have never departed from the fundamental common law principle that contracting parties must be able to access a neutral arbiter; they have merely expanded their horizons as to who can fill this role. And certainly there is nothing in the Arbitration Act that requires courts to embrace arbitration at all costs – section 7(2) expressly provides maintains the court’s historic power to declare an arbitration agreement invalid.
There is one further potential reason Brown J.’s concurrence is correct, which none of the justices discuss, including Brown J. Section 7(2) of the Arbitration Act states that a court may refuse to grant a stay of proceedings only where the arbitration agreement is invalid. However, as the doctrine of unconscionability is an equitable doctrine, it does not concern the validity of the agreement, only its enforceability. The law has long distinguished between these two concepts, validity speaking to the essential elements of the contract and enforceability concerning whether the agreement will be enforced in a court of law. A valid agreement (ie, one in which there has been offer and acceptance, with good consideration, between parties with capacity who consent, and for a legal purpose) may still be unenforceable for a variety of reasons. The implication of the majority’s decision is that the arbitration agreement was in fact valid, but that it was not enforceable due to its unconscionability. Conversely, the doctrine of public policy is generally understood to concern the basic legality of the agreement, and thus its validity.
To be sure, the courts have been known to speak of validity and enforceability interchangeably, so there is an arguable case that the term “invalid” in the Arbitration Act should be read broadly to include arbitration agreements that are unenforceable. But, by the same token, the terms remains legally distinct. As late as 2017, in Douez v. Facebook, Inc., the plurality for the Supreme Court drew a distinction between the “formal validity of the contract” and whether the court should enforce it (see para. 33). Section 7(2) could certainly have been drafted to include the term “unenforceable” and its omission arguably indicates that a stay will only be refused if the arbitration agreement is invalid as opposed to merely unenforceable. If this is the case, a stay could never be refused on the basis of unconscionability, but it could be refused based on the doctrine of public policy.
What I find truly curious about the Uber decision is that none of the other justices joined Brown J.’s concurrence. Remarkably, the majority decision does not even mention the public policy aspect of the concurrence. It may be that the justices in the majority felt uncomfortable going down this road since the parties did not argue the appeal on the basis of public policy. Whatever the case may be, the failure to address and rebut Brown J.’s argument may ultimately prove to be the decision’s silver lining. Future lower court judges will now be free to adopt Brown J.’s reasoning on the public policy doctrine and apply it where circumstances permit. Moreover, I suspect many judges will pay heed to his criticisms of the majority’s approach and will hesitate to apply the unconscionability doctrine. His well-reasoned and persuasively argued concurrence will serve as wisdom and warning to future panels. Like a warning label on a vial of medication that reads CAUTION in bold uppercase letters, Brown J.’s concurrence tells lower court judges that this precedent comes with complications and should not be consumed unless absolutely necessary.
The future is anyone’s guess, but I will not be surprised if lower courts continue to cite Brown J.’s concurrence and a majority of the Supreme Court eventually affirms it. Doctrines that are principled, coherently articulated and that can be consistently applied have a tendency to prove themselves more durable over the long haul. Results-oriented or highly discretionary doctrines, by contrast, can be frustrating for judges to apply and invariably create a legal morass. Where doctrines are not applied in a predictable and consistent fashion, they often become unworkable and eventually need to be replaced. We have seen this phenomenon play out in the context of administrative law time and again (though hopefully the recent decision in Vavilov represents an end to this trend). But in the realm of private law, the doctrine has generally developed in a much more thoughtful and incremental manner. And for this reason, I suspect that the judiciary will eventually turn to Brown J.’s concurrence, both to navigate the doctrine of public policy and to re-stabilize the doctrine of unconscionability and contract law more broadly.