Uber v. Heller: Class Actions in the Crossfire | July 6, 2020

Print Friendly, PDF & Email

Simon V. Potter, Ad. E.1 | potter@simonpotter.ca | +1 (514) 919-4268

In the immediate wake of the June 26, 2020, judgment of the the Supreme Court of Canada in Uber Technologies v. Heller2 (“Uber”) considerable attention has been paid to what this judgment means for arbitration, for arbitration clauses, and for what has so far been the governing rule of primary deference to arbitrators on the question of their own jurisdiction.

The judgment is grounded, for seven of the nine judges, on an expanded view of unconscionability3 and of the circumstances in which Canadian courts can set aside “improvident bargains”, and it may be that this also will generate uncertainty in a variety of areas outside the context of arbitration clauses.

The judgment is rendered, though, in the context of a putative class action and we should be paying attention also to what a sizeable majority has just said about class actions in Canada.4 This article explores what may turn out to be a series of consequences and newly created uncertainties, at least some of them unintended but nevertheless inescapable.

In summary, it is likely that the majority, in its desire to react to “an arbitration agreement which makes it impossible for one party to arbitrate”, which it saw as a “classic case of unconscionability”5, has made a variety of assumptions, some implicit but many explicit, which will now challenge many of Canada’s carefully nourished views of what a Canadian class action is, how it operates, and how it comes to bind not only defendants and a representative plaintiff but, as well, a whole class of plaintiffs.

The Class Action of Mr. Heller

This class action was first proposed by proceedings filed in 2017. Mr. Heller makes a variety of claims all of which the Uber majority sees as hinging on applicability of Ontario’s Employment Standards Act, 2000, S.O. 2000 c. 41 (“ESA”). “The essence of Mr. Heller’s position is that he is an employee within the meaning of the ESA.6

Faced with an arbitration agreement between the parties, the courts and eventually the Supreme Court of Canada had to decide whether the claim should be handled by Ontario’s courts or by the designated arbitration process.

Prejudging the issues of the class action

A decision as to whether a particular contractual clause, here a clause by which the parties agree to refer their disputes to mediation and arbitration, is unconscionable does not necessarily require that a decision first be made as to whether the contract or the dispute7 is essentially commercial8 or not, or as to whether they are “fundamentally about labour and employment” or about an independent contractor signing on for piecework.

Even if the class action had reached the stage of certification or authorization of the class action, a court, including the Supreme Court of Canada, would certainly have taken pains to leave this decision to the judge deciding which issues to certify and for which definition of class, and then to the judge hearing the post-certification action on the merits.

Here, though, even though at a stage well before certification, indeed at a stage at which it was decided that the application for certification should not be stayed but allowed to move forward, the majority has allowed itself to prejudge these issues, at least in part.9 Both the issue as to the categorization of the contractual relationship, at the very heart of the putative action, and the issue whether any Uber driver in Ontario is bound to respect his or her promise to submit to arbitration.

Though the majority properly says10 that it is careful to look only at the dispute itself, not at the actual factual nature of the contractual relationship, to come to the conclusion that what is at stake is whether Mr. Heller is an employee11, it has essentially decided that the dispute is therefore not about whether Mr. Heller (and presumably the other members of the proposed class) signed on to become an independent contractor, as the contract actually says12. If it were about that, the majority’s own logic would have the ICAA apply.13

The purpose of this article is not to say that either of these conclusions was necessarily wrong, but it is to point out that we have witnessed, in a matter having to do principally with choosing whether to have a dispute settled by arbitration or by domestic class action, a pre-certification prejudging of how one of the issues central to the class action should eventually be decided.

It is likely that the majority had in mind to call this dispute non-commercial out of a desire to avoid its extraction of Mr. Heller from arbitration having wider repercussions in a clearly commercial context14, but there are two things to say about this. Firstly, it is not only in non-commercial matters that an arbitration clause can be found in an adhesion contract, or in a contract in which one party has a greater bargaining power than the other: unconscionability can be found in commercial contexts, too.15 Secondly, and more importantly for the purposes of this article, even to posit that the majority sought to limit the downstream ramifications of its decision by camping it in the narrow sphere of uncommerciality is recognition that the majority did in effect make the finding.

The normal order of things would have been to allow a certification judge to decide whether the action presented essentially an employment law question or a commercial issue as to an independent contractor’s rights and recourses as a licensee of a computer program. Class counsel can surely now be counted on to argue to the Ontario courts, pre- and post-certification, that it has already been decided that the ESA applies.

Even if the Ontario courts muster the courage to say that it has not been decided, the class has already won one issue and Uber has lost it, on a class-wide basis: the issue whether Mr. Heller or any member of the class is bound by the arbitration clause twice16 agreed to.

Presuming that certification or authorization should be granted

Indeed, it is difficult to conclude otherwise than that the majority has not only instructed lower courts to rule this way on this central issue if ever the class action is certified, but that it has also taken for granted that there will be certification.

A first-instance judge will now be extremely hard-pressed to decide that Mr. Heller’s claim should not be certified, or should be allowed to proceed only as an individual claim17, or that he should be left to revert to the unconscionable arbitration.

That is, the majority appears clearly to have presumed, pre-certification, that there should be a class action to determine Mr. Heller’s claim.

Presuming that Mr. Heller’s recourse must be by a class action before Ontario courts

On several occasions, the majority examines what is seen as the high cost of Mr. Heller’s pursuing his individual claim before the contractually imposed ICC, and the cumulative of costs of other Ontario Uber drivers also filing their arbitration claims, to the presumably less unconscionable cost of pursuing a class action in Ontario.

The concurring judgment of Brown J. is similar to the extent that he sees the matter as one of access to justice, the cost of the ICC arbitration, in particular the up-front filing fee, being so prohibitive that it amounts to a denial of access to justice. His conclusion is also a factual one, in that he finds that the clause makes the “enforcement of law practically impossible”, if not because it “expressly blocks access to a resolution” at least because it “has the ultimate effect of doing so”.18

These conclusions are reached without any consideration19 of

a) Any arbitrator’s power to make awards, including interim awards, in relation to costs, including filing fees;

b) The likely comparative costs of running a class action which puts at stake hundreds of millions of dollars20 of a carefully thought-out business plan;

c) The availability to Mr. Heller of the benefits of contingency fee arrangements generally applicable in class actions and certainly available in arbitrations;

d) The availability of third-party funding whether of the arbitration or of the class action;

e) Whether in light of these solutions for litigation financing, the filing fee was indeed an “insurmountable precondition (or) total barrier to court access”, either in absolute terms or in comparison to the costs and other barriers to the bringing of a class action, or whether the fee might as well have been a “brick wall” just as would have been “an upfront payment of 10 billion dollars”;21

f) The possibility of a class action involving Mr. Heller’s claims being handled by way of an arbitration;22

g) The fact that most arbitration institutions, including the ICC, have adopted simplified, accelerated23 and cheaper procedures for lower-value claims;

h) The fact that arbitrators have discretion to make procedural orders required to hasten and simplify the cases before them, and to avoid their becoming disproportionately costly;24

i) Whether the filing fee is “warranted in light of the parties’ relationship and the timely resolution that arbitration can provide”25; and

j) The fact that an arbitrator, even if the seat of the arbitration is Amsterdam (as is the case here26), can hold hearings wherever appears the most convenient and proportional.

Now, it may be that exploring these aspects of a finding of unconscionability, or of effective denial of access to justice, would have sufficiently complicated the issue of jurisdiction to require “referring the question to the arbitrator out of respect for the competence-competence principle”27. It does seem to the reader that the majority avoided anything which would exceed “only superficial consideration of the documentary evidence in the record”28. But this only indicates that the majority found a “superficial consideration”29 sufficient to warrant an essentially factual finding that it would be unconscionable to require proceeding as the parties had agreed rather than as Ontario class action law permits.

In dissent, Côté J. writes that the majority could not decide what it did “without usurping the role of the arbitral tribunal”. It must also be that the majority has usurped the role of the certifying judge.

A finding of unconscionability going well beyond the arbitration clause

Though the immediate question before the Court was whether to defer to the arbitrator according to the arbitration agreement, or to permit the representative plaintiff to proceed through the Ontario courts, and though the issue of unconscionability was decided by the majority in large part on the basis of what was seen as disproportionate burden arising from the arbitration clause, the unconscionability was also found on parameters which will surely go the class action court’s assessment of the overall contract at issue.

The majority has found that there was, as a matter of fact, an inequality of bargaining power at the time of contract. This is one of two criteria which must be met, according the majority30, for a finding of unconscionability. It is not conceivable that this inequality of bargaining power would be found to have existed in relation to the arbitration clause but not to all the other clauses of the adhesion contract. As the majority wrote, with its own emphasis: “Unconscionability, in our view, is meant to protect those who are vulnerable in the contracting process from loss or improvidence to that party in the bargain that was made.”31

The Court’s decision of Mr. Heller’s comparative weakness in the contracting process lies exclusively on factors which would apply to his whole class.32 The contract was standard form, a contract of adhesion. There is a “significant gulf in sophistication between” an Uber driver and “a large multinational corporation”. It would be a “rare fellow” who would read the whole contract or, with that reading, “suspect that behind an innocuous reference to mediation … followed by arbitration … there lay a US$14,500 hurdle to relief.”

The only other criterion of unconscionability is that the bargain be found to have been “per se unreasonable”33 or, simply, “improvident”.34 That improvidence can be found in any clause which turns out unduly to advantage or enrich the presumably stronger party35 when seen in the context at the time of contracting.36 It is not an exercise of “exact science” but “inherently contextual”.37 That said, the US$14,500 filing fee is found to be a “clear” “improvidence”.38

Further, finding the manifestly unfair bargain can be enough for the court to find that it must have resulted from unequal bargaining power.39 Criterion number two is itself the answer to criterion number one40, and there is no need to find that the unfairness or the improvidence is “gross” or that it results from an intention to take undue advantage.41

Brown J. is surely on solid ground to write that the majority has concluded “by vastly expanding the scope of the (unconscionability) doctrine’s application and removing any meaningful restraint” and “drastically expanding the doctrine’s reach without providing any meaningful guidance as to its application”42, and that “ their approach is likely to introduce added uncertainty in the enforcement of contracts, where predictability is paramount.”43

This will be so whether in the presence of an arbitration clause or not, and whether the issue arises in the context of a class action or not44, but it is clear that the expansion of the doctrine here, in a class action, opens the door to many more disputes pre-certification and, as here, to more pre-certification judicial determinations of fact.

It is impossible in all this light to conclude otherwise but that the majority has decided, prior to certification or discovery or trial, that the inequality of arms and the improvidence existed for all the drivers who twice clicked “I agree” to the contract with Uber. Though the majority stresses that the unconscionability can be found in relation to the arbitration clause, separately from the contract as a whole,45 half the “duality”46 will now surely be considered settled fact in the yet to be certified class action and the other half can be met with a showing of undue advantage or enrichment. Brown J. fears not that it is unlikely that the finding of unconscionability can be constrained only to the arbitration clause but that it is in law impossible: “Unlike public policy considerations that target a specific contractual provision, unconscionability’s substantive inquiry must consider the entire bargain…”47

Legally permissible or not, effectively constrained or not to the arbitration clause, the majority’s embarking on this “substantive inquiry” and deciding it on the basis of a superficial review of whatever few facts appear on the record, runs counter to many decades of jurisprudence to the effect that the first-instance court is a gatekeeper essential to the integrity of the class action and, then, an essential fact finder on the merits.

An openness to other pre-certification determinations

The majority decided that it was not necessary to rule whether the arbitration agreement between Uber and Mr. Heller was void as having “the effect of contracting out of mandatory protections in the ESA”, but only because the majority had concluded the arbitration agreement was invalid because unconscionable. 48

This is an indication that the majority, had it not reached a whole-class determination that the arbitration clause was unconscionable and not binding, would have proceeded to decide whether the arbitration agreement illegally sidestepped mandatory protections in the ESA and was therefore void. That is, if the majority had not stepped into the arbitrator’s competence to decide his or her own jurisdiction one way, it would have done it another way.

This issue, whether the agreement does or does not seek to avoid, or succeed in avoiding, statutory requirements mandatory in an Ontario employment context, is complex and factually dense. Deciding that issue would not only involve preempting the arbitrator as to the effect of contractually agreed choice of law provisions and as to an arbitrator’s power49 to apply mandatory provisions of any particular jurisdiction notwithstanding the general choice of law, but also require pre-judging the question whether the contract truly does impose results incompatible with the Ontario statute.

This would necessarily demand coming to a landing on the issue whether Mr. Heller is, whatever his contract says, an employee within the meaning of that statute. This is not only a factually laden inquiry, but is also an inquiry leading to one of the very conclusions Mr. Heller seeks on the merits.

It is noteworthy that the majority did not decline to embark on this inquiry because it would have been incompatible with what we have until now understood a class action to be, but because another solution was found which itself puts into question our understandings about the proper working of a class action, of class action certification, and of trial and eventual class-wide judgment.

The broad swath

The majority took pains to narrow the ambit of its finding here, stressing what it found to be the non-commercial nature of the dispute before it, and stressing that this should be considered one of those rare, “abnormal” cases in which deference to the arbitrator should not be the rule.

However, any class action plaintiff, whether raising a commercial dispute or not, will now have the benefit of an additional argument on which to argue invalidity of a contract and to do so on a class-wide basis.

Class action plaintiffs will clearly have the benefit of at least bringing to court actions which would normally have been left to arbitration, waiting for the application to suspend the action in favour of the arbitration proceedings, and then arguing that the arbitration clause should not apply.

Even in areas not having to do with arbitration, those plaintiffs will have additional grounds on which to raise matters of unconscionability, and perhaps even to get rulings (as to at least one of the dual criteria of unconscionability) before certification, let alone before trial. If Brown J. is right, that the majority has paved the road towards judgments now being rendered on the basis of “unreasoned intuition and ad hoc judicial moralism”50, class action plaintiffs will feel that they have little to lose in taking pre-certification or pre-trial chances at seeing which way the wind blows.

Any class action relying on, or seeking to avoid, an adhesion contract51, whether it be in a commercial matter or not, whether with individuals as signatories or corporations, will be especially susceptible to plaintiff strategies such as these.

Defendants in class actions, concomitantly, have reason to think twice before invoking arbitration clauses or indeed any other agreement to avoid normal recourse to the courts, for fear that doing so will result in pre-certification findings that the agreement is partially or even wholly “unconscionable”.


Many (not all, but many) of these troubling uncertainties for the future would have been avoided by the majority’s adopting the approach of Brown J., clearly and forcefully designed to dissuade drafters of adhesion arbitration agreements from loading the dice52, but we are faced with a judgment of seven out of nine judges which will have consequences for much more than arbitration agreements and much more than the competence-competence principle.

Their judgment will affect the enforcement of civil and commercial contracts, and will, for good or ill, affect the way in which class actions reach certification and, eventually, trial.


1 The author practises as an arbitrator and mediator at www.simonpotter.ca, and offers advice as well in, among others, matters of commercial litigation and class actions, fields in which he has many years of experience. He was a member of the Canadian Bar Association’s Task Force on Multi-jurisdictional Class Actions.
2 2020 SCC 16, https://decisions.scc-csc.ca/scc-csc/scc-csc/en/item/18406/index.do .
3 As was the decision of the Ontario Court of Appeal: 2019 ONCA 1, 430 D.L.R (4th) 410.
4 Brown J. concurs in the result, declaring invalid the mandatory arbitration clause, but on the ground that it undermines the rule of law by in fact foreclosing the very access to legally determined dispute resolution which it advertises. Uber, 101-102.
5 This, advanced by the authors Abella and Rowe JJ. at paragraph 4 of the majority judgment, is perhaps a non-sequitur, and is certainly an equation which Justice Brown, concurring in the result, preferred to avoid by seeing (though this point had not been advanced by Mr. Heller) a public-order-violating deprivation of access to justice: Uber 13, 106.
6 Uber, 1, 12. 
7 Uber, 25.
8 As a general matter or as the term “commercial” should be understood within the context of the International Commercial Arbitration Act, 2017, S.O. 2017, c. 2, Sch. 5 (“ICAA”), or within that of the UNCITRAL Model Law on International Commercial Arbitration, U.N. Doc. A/40/17, Ann. 1, or under any of its footnotes. See Uber, 14, 16, 19, 21, 23.
9 Uber, 19.
10 Uber, 25.
11 Uber, 26.
12 Uber, 216, per Côté J. “The Service Agreement expressly states that it does not create an employment relationship. Instead, it is a software licensing agreement, which … is a type of transaction that is identified as coming within the scope of the UNCITRAL Model Law.”
13 See Uber, 23 and 24, and references there to language tending strongly in favour of this conclusion.
14 The majority was explicit in calling this an “abnormal” instance where that extraction was necessary, but that was because of the perceived unconscionability, not because of the non-commercial nature of the dispute: Uber, 37.
15 As the majority recognizes (Uber, 55).
16 Uber, 7.
17 There is not a line in the majority assessing the difficulty for Mr. Heller to pursue his Ontario claim as an individual rather than as representative of a class.
18 Uber, 113.
19 Côté J., in her dissent, found it impossible to decide the question of unconscionability or the applicability of the ESA on the basis of a superficial inquiry. Uber, 201, 217.
20 Mr. Heller’s claim is for CAN$400,000,000 (Uber, 188). An arbitration involving such a claim would see total costs of considerably more than US$14,500. A class action filed in Ontario would cost many, many times that.
21 Uber, 114.
22 The ICC rules do not yet entertain this, but other arbitration bodies do handle class actions, and even an ICC ruling on an individual case (or several combined for hearing) can result in a precedent for other claims.
23 The class action proceeding was three years old by the time of the Uber pre-certification decision of the Supreme Court of Canada. It will presumably have many years to go before it reaches even a first-instance decision on the merits. Arbitrations exist in part because they are mercifully more quick.
24 The majority writes (42), for example, that “Courts have many ways of preventing the misuse of court processes for improper ends”, but never considers that arbitrators also have many tools and discretionary powers to avoid one party abusing the arbitral process.
25 Uber, 130.
26 Uber, 8.
27 Uber, 31.
28 Uber, 32, in the majority’s discussion of the applicability of the criteria set out in Dell Computer Corp. V. Union des consommateurs, [2007] 2 S.C.R. 801, and Seidel v. TELUS Communications Inc., [2011] 1 S.C.R. 531.
29 Uber, 32-34, with reference also to Rogers Wireless Inc. v. Muroff, [2007] 2 S.C.R. 921.
30 Uber, 63.
31 Uber, 60, 66.
32 They have nothing to do, for example, with subjective and individual matters such as “the claimant’s ‘purely cognitive, deliberative or informational capabilities and opportunities’” or with the applicant driver being “seriously volitionally impaired or desperately needy” (Uber, 67) or with his “ability to understand or appreciate the significance of the contractual terms” (Uber, 68). All of the grounds advanced by Mr. Heller have to do with issues which any Ontario Uber driver might raise (see Côté J. at 28).
33 Uber, 63.
34 Uber, 60, 62, 64, 65, 73, 79.
35 Uber, 74, 76.
36 Uber, 75.
37 Uber, 78.
38 Uber, 94.
39 Uber, 79.
40 Contra, per Brown J.: Uber 156.
41 Uber 80-85. Contra, per Brown J.: Uber 160-166.
42 Uber, 103.
43 Uber, 147.
44 As Brown J. writes, “to introduce uncertainty to the enforcement of contracts generally”. Uber, 103, 154, 170.
45 Uber, 96.
46 Uber, 63.
47 Uber, 172.
48 Uber, 99.
49 Duty, according to many authorities in the field of arbitration.
50 Uber, 153.
51 Côté, in dissent, writes (Uber 257, 266), “In particular, I am concerned that their threshold for a finding of inequality of bargaining power has been set so low as to be practically meaningless in the case of standard form contracts.” “I fear that the effect of their approach amounts to a sweeping restriction on arbitration clauses in standard form contracts ….”
52 A lesson which those drafters should take to heart, and which arbitration practitioners should also deliver if they have at heart the integrity of the arbitration process as a legitimate alternative method of resolving disputes. But that is not the point of this article.
Services en conseil et règlement de litiges

Simon V. Potter, Ad. E.

Plus de quatre décennies de pratique juridique dans les litiges commerciaux et constitutionnels complexes, l’arbitrage et la médiation, le droit administratif, le droit commercial et le droit international. Reconnu à plusieurs reprises comme un leader dans ces domaines…    Lire la biographie → 


— Tél: +1 (514) 919-4268

— Courriel: potter@simonpotter.ca