Like a poor Marksman, ONCA keeps missing the Arbitration target
By: Anthony Daimsis*
This comment calls out the ONCA’s struggles with foundational arbitration principles like seat theory, competence-competence, and uniformity with respect to international arbitral legal instruments. The ONCA’s struggles risk exposing Ontario as an undesirable arbitral seat. Any good arbitration’s protagonist is the tribunal, but even the best tribunal needs a good supporting cast in the form of supervisory courts. Ontario courts must adapt to fit the part.
This comment discusses two recent cases from the Ontario Court of Appeal (“ONCA”) and one from the BVI Commercial Court (Eastern Caribbean Supreme Court, Commercial Division, Territory of the Virgin Islands) (“BVI”). The judge in BVI, Justice Barry Leon, is a Canadian who spent his whole career in Canada, hence why I incorporate this decision.
The two ONCA cases examined in this comment include The Walt Disney Company v. American International Reinsurance Company, Ltd. (“Disney”)  and Heller v. Uber Technologies Inc. (“Uber”). The BVI decision is Hualon Corporation (M) Sdn Bhd (in receivership) Acting by its Receiver and Manager Mr. Duar Tuan Kiat v. Marty Limited (“Hualon”).
In each case, the court confronted thorny questions of arbitration law, however the BVI court confronted them best by adopting an approach aligned with international practice. Particularly, the BVI court accounted for the relevant legislation’s international origin and the need to promote uniformity in its application, an approach that judges in Ontario could learn from. Had the ONCA surveyed other jurisdictions, it would have found copious cases and authorities clarifying the legal effects of choosing a seat, of competence-competence and of uniform interpretation and application.
For example, in the ONCA’s Disney endorsement, the Court missed an opportunity to learn from other jurisdictions about an important arbitration doctrine: seat theory. In its Uber judgement, the Court missed an opportunity to sort through important differences between domestic and international arbitration and the competence-competence doctrine.
In contrast, in Hualon, the BVI Court decided in line with what arbitration specialists expect: a decision harmonious with other modern arbitration jurisdictions. This note argues that it is this approach Ontario courts should emulate to confirm Canada as a leading arbitral centre.
I will begin with some context on the lower court’s decision. The Ontario Superior Court of Justice faced a tricky question made trickier owing to the parties’ unnecessarily complicated arbitration agreement. But to put it simply, the question was whether Disney had properly commenced arbitration in line with the parties’ agreement.
Disney believed it had and sought the court’s assistance to appoint an arbitrator, arguing that the parties had not managed to agree on an arbitrator. American Reinsurance argued that, because Disney had not followed the procedures agreed to in the arbitration agreement, no failure to appoint an arbitrator occurred because arbitration had not properly commenced.
Madame Justice Pollack agreed with American Reinsurance, finding that Disney relied on a procedure to which the parties had not agreed, and thus Disney had neither properly commenced arbitration proceedings nor followed the proper method to appoint an arbitrator.
The Arbitration Agreement
The parties’ arbitration agreement offered several locations to seat their arbitration: Hamilton, Bermuda; London, England; or Toronto or Vancouver, Canada. If the Parties selected a location in England or Canada, they agreed that the English Arbitration Act 1996 (EAA) would govern their arbitration.
The clause is worth a look, if only to marvel at the unnecessary complications:
- It is agreed that any dispute arising out of or in connection with this policy, including any question regarding its existence, validity or termination, shall be referred to and finally resolved solely by Arbitration or Alternative Dispute Resolution (ADR).
- This policy shall follow the terms of the Followed Policy with respect to Arbitration or ADR only when said terms require the dispute to proceed either in Hamilton, Bermuda or London, England or Toronto, Canada or Vancouver, Canada.
- If paragraph F.2 above does not apply, the Insured shall select either of the following venues and procedural laws for such Arbitration or ADR:
(i) Arbitration held in Hamilton, Bermuda under The Bermuda International Conciliation and Arbitration Act of 1993 (exclusive of the Conciliation Part of such act), as may be amended and supplemented, which Act is deemed incorporated by reference into this clause, or
(ii) Arbitration held in London, England, Toronto, Canada or Vancouver, Canada under the English Arbitration Act of 1996, as may be amended and supplemented, which Act is deemed incorporated by reference into this clause.
Disney selected Toronto.
By electing to seat their arbitration in Toronto, the parties’ arbitration became subject to Ontario’s mandatory provisions found in Ontario’s International Arbitration Act (ICAA), supplemented by the EAA provisions – not as law but as party stipulations. Consequently, Ontario’s mandatory arbitration provisions, found in the ICAA, applied, with ICAA’s non-mandatory provisions replaced with equivalent provisions from the EAA. Complicated, but workable. The visual below may help sort through this clause:
Did Disney Properly Commence Arbitration?
In finding that Disney had not properly commenced arbitration proceedings, Justice Pollack found that the parties agreed to apply the procedures “of the English Arbitration Act, with respect to appointing arbitrators”. As explained below, her finding was incorrect. Nevertheless, from this finding, Justice Pollack concluded that because Disney used a procedure different from the EAA, it did not comply with the arbitration agreement and thus did not properly request arbitration. Consequently, Justice Pollack rejected Disney’s request for assistance appointing an arbitrator.
Why Justice Pollack was wrong
What is relevant in Disney is that EAA section 14 governs arbitration commencement, not because the EAA is law but because ICAA’s equivalent provision (Model Law Article 21) is non-mandatory. Hence, the parties were free to opt out from this non-mandatory provision and replace it with their stipulated EAA provision. This subtle point is critical but by all appearances both Justice Pollack and ONCA missed it.
The added twist making this case even trickier – is that EAA section 14 provides that arbitrations commence once the arbitral tribunal appointment procedure is under way, while the ICCA dates the commencement proceedings to the respondent’s receipt of the notice of dispute. Furthermore, under the ICCA (and as here, where the Parties chose Toronto as their seat), if parties fail to appoint an arbitrator, the Ontario Superior Court steps in to appoint one. This is because ICAA’s provision on appointing arbitrators, specifically Model Law Article 11(4), is mandatory. As a last resort, the article provides, it is the court at the seat (here, Toronto) that finally appoints the arbitrator.
This matters because when Justice Pollack determined that Disney was bound to use the EAA to finally appoint arbitrators she was wrong. Her error led her to find that because Disney had not followed the EAA’s procedure, it had not properly attempted to commence arbitration. Justice Pollack therefore decided that Ontario courts could not help Disney. But this ignores the parties’ seat selection – Toronto, meaning that the ICCA’s mandatory provisions governed regardless of what the parties may have wanted. In fact, Ontario courts should step in to appoint an arbitrator when the parties’ chosen method fails, as it did here.
Consequently, since an Ontario court had to step in to appoint an arbitrator and since the EAA deems an arbitration to commence once the procedure to appoint arbitrators is under way, the parties’ arbitration should have commenced once an Ontario court properly stepped in to appoint an arbitrator. Justice Pollack short-circuited all of this by misunderstanding one of arbitration’s keystone ideas: seat theory.
Seat theory holds that parties may not avoid the seat’s mandatory arbitration laws; consequently, these mandatory provisions apply regardless of what the parties may agree on. This territorial approach—which endorses international law’s general principle that a state is sovereign within its own borders—confirms that states may regulate proceedings occurring within their territory. While the Model Law endorses party autonomy (Model Law Article 19), this endorsement is circumscribed through its mandatory provisions. For this case, this means respecting the parties’ choice to include the EAA, but subordinating it to the ICCA because Ontario was the seat of arbitration.
ONCA’s lack of commercial sense
Despite seat theory’s prominence as one of arbitration’s foundational theories, the ONCA paid it no heed. In disagreeing with Disney, the court stated that the “interpretation advanced by the appellant does not make commercial sense because it would mean that different procedural laws would apply depending on whether the arbitration took place in Toronto or Vancouver.” In fact, seat theory means exactly this.
British Columbia’s mandatory arbitration provisions govern an arbitration conducted in Vancouver; and these provisions may differ from those found in Ontario’s legislation. Hence, a different seat or venue can mean different procedures. That’s the point. But one need not know arbitration’s inner workings to realize this. Canadian constitutional law reminds us that arbitration falls within sections 92(13) and (14) of the Constitution Act 1867. Hence, each province is free to adopt rules on arbitration, which may differ from those of other provinces.
Perhaps even more astonishing is the very next sentence in the ONCA’s endorsement: “Moreover, it makes no sense that the parties would agree to have an Ontario court apply specific provisions of the [English Arbitration] Act relating to the matters set out in s. 2(2), (3) of the Act.”
Sections 2(2) & (3) of the EAA, which find their equivalent in Model Law Article 1(2), make the practical point that regardless of where an arbitration takes place, if a party needs to compel a witness within that court’s jurisdiction, freeze assets found within that court’s jurisdiction, or other similar matters, a court in that relevant jurisdiction is competent, regardless of the parties’ chosen seat. Such provisions are vital to (and found ubiquitously in) modern arbitration legislation.
These provisions were designed so courts could support arbitrations taking place within their jurisdiction. And this is exactly what ONCA thought made no sense. The parties chose the EAA to govern their arbitral procedures and therefore agreed that courts, in any jurisdiction and to the extent that their rules allow, may issue measures contemplated under section 2.
The ONCA’s reasons are far from what practitioners expect from Ontario’s highest court, this risks signalling to the outside world that Ontario’s highest court is ill-prepared to support arbitration.
The UBER decision
In Uber Technologies v Heller, the ONCA per Nordheimer JA declared an arbitration agreement between Uber and one of its drivers invalid. The consequence of his decision is serious enough that the Supreme Court of Canada has accepted to hear Uber’s appeal.
Brief factual overview
Mr. Heller, on behalf of Uber drivers in Ontario, commenced a class action against Uber for what he deemed breaches of the Employment Standards Act, 2000, S.O. 2000, c. 41 (the “ESA”). Additionally, he asked the court to declare void and unenforceable all arbitration agreements in contracts between Uber and its drivers in Ontario. The action claimed damages of $400 million.
The lower court denied Mr. Heller his relief but the ONCA reversed the lower court’s decision and declared the arbitration agreement between Uber and its drivers void on grounds of unconscionability. Nordheimer JA relied exclusively on domestic law, ignoring the choice of law clause naming Dutch law, and neutering the notion of shared jurisdiction between arbitrators and judges to enforce arbitration agreements. This section explains where the ONCA went wrong.
Domestic or International – a lot turns on this distinction
The ONCA adopted the lower court’s view that nothing in the case turned on whether the ICAA or Ontario’s domestic Arbitration Act, 1991 applied to the arbitration clause in issue. This seemingly innocuous finding set off an unfortunate cascade of errors.
Justice Nordheimer’s election to apply the domestic act without considering which act should apply hampered his decision from the outset. The two acts use different interpretive rules, which can lead to differing approaches to competence-competence.
Justice Nordheimer applied the domestic act without any legal rationale. This matters because unlike the domestic act, the international act requires interpretation based on the international act’s origin and the desire of its drafters to promote uniformity in its application.
Model Law article 2A(1), puts into writing what most seasoned practitioners already know: interpret the Model Law bearing in mind its international nature and the need to promote uniformity. Concretely, this means decision-makers must recognise the Model Law’s own interpretation rule, distinct from domestic legislation. Interpreting the Model Law, decision-makers should also look to the Model Law’s travaux (as the ICCA suggests at its s. 6(3)) to understand its drafters’ intentions in any article’s wording and phrasing. Additionally, decision-makers should examine how other Model Law jurisdictions interpret and apply the Model Law. These methods reinforce the uniformity mandate entrenched in 2A.
This leads to consistency amongst Model Law jurisdictions by ensuring, for example, that a court interpreting the same arbitration agreement in any Model Law jurisdiction will reach the same conclusion about its enforceability. As explained below, Dutch law governed this arbitration agreement. Hence, Nordheimer JA should have assessed its unconscionability using Dutch law. But he didn’t.
As explained below, one of Justice Nordheimer’s errors was to analyze unconscionability strictly from a Canadian perspective. But equally troubling is how he misapplied competence-competence.
Had Justice Nordheimer looked to other jurisdictions he may have avoided misunderstanding competence-competence.
According to Justice Nordheimer, competence-competence only “addresses situations where the scope of the arbitration is at issue”. The issue of validity, in his view, “is one for the court to determine as s. 7(2) of the Arbitration Act, 1991 makes clear”. But s. 7(2) does not say this. Instead, like the ICCA, the domestic act adopts a shared jurisdiction allowing both courts and arbitral tribunals to assess the validity and existence of arbitration agreements, as the relevant provisions show.
|Domestic Arbitration Act||International Arbitration Act (ICCA)|
7 (1) If a party to an arbitration agreement commences a proceeding in respect of a matter to be submitted to arbitration under the agreement, the court in which the proceeding is commenced shall, on the motion of another party to the arbitration agreement, stay the proceeding.
(2) However, the court may refuse to stay the proceeding in any of the following cases:
2. The arbitration agreement is invalid.
17 (1) An arbitral tribunal may rule on its own jurisdiction to conduct the arbitration and may in that connection rule on objections with respect to the existence or validity of the arbitration agreement.
8 (1) A court before which an action is brought in a matter which is the subject of an arbitration agreement shall, if a party so requests not later than when submitting his first statement on the substance of the dispute, refer the parties to arbitration unless it finds that the agreement is null and void, inoperative or incapable of being performed.
16 (1) The arbitral tribunal may rule on its own jurisdiction, including any objections with respect to the existence or validity of the arbitration agreement.
Shared Jurisdiction between courts and arbitral tribunals
Without creating a hierarchy between courts and tribunals, both acts deliberately endorse a shared jurisdiction between tribunals and courts. Sections 7 and 17 of the domestic act and articles 8 and 16 of the international act reveal that both courts and tribunals may address questions about an arbitration agreement’s existence or validity. What the acts take no position on is who should first answer the question. To this, the Supreme Court of Canada provided the answer, at least for domestic arbitration.
When courts must defer to arbitral tribunals
In Dell Computer Corp. v. Union des consommateurs (“Dell”), The Supreme Court of Canada offered a general rule that jurisdictional objections (including unconscionability in Uber) should first go to the arbitral tribunal. The Supreme Court tempered this general rule with an exception tied to the type of objection the party raised: if the objection is a pure question of law, the court could first address the question. If the objection is a question of fact, the tribunal should first address the question. If the objection is a question of mixed law and facts, the tribunal should first answer the question, unless the facts require only a superficial investigation, in which case the court could first address the question. In explaining why the Court should refer cases to arbitration in the face of factual and mixed facts and law, it joined a chorus of other countries’ high courts that share the modern view that the court should normally refer the case to arbitration, as arbitrators have, for this purpose, the same resources and expertise as courts. However, Dell is not the approach favoured internationally, if at all.
Dutch law governed unconscionability
Generally, unconscionability requires investigating beyond superficial facts. In discussing whether Uber drivers are employees, Perell J. considered this “is a fact-based determination that depends upon on (sic) a variety of factors and not just the written or oral agreement between the parties.” At minimum then, the ONCA should have considered whether it could address the question superficially or not.
If it had, relying on Dell, the ONCA should have referred the question of the arbitration agreement’s validity to the arbitral tribunal since the question needed a thorough assessment of facts as well as law.
Further troubling were the facts on which Justice Nordheimer relied to support his unconscionability analysis. He failed to recognize that Uber drivers could have worked for other food delivery services similar to Uber Eats that do not require the type of expensive dispute resolution Justice Nordheimer found so egregious. Skip the Dishes, for example, in particular clause 12 of its terms, calls for arbitration wherever is convenient in Ontario, using ADR Institute of Canada (located in Toronto) and applying Ontario law. Skip the Dishes is a service very similar, if not identical to Uber Eats, nor is it the only other player in the market.
But most importantly, Justice Nordheimer ignored the Supreme Court’s directive to allow the arbitral tribunal to first address the question. But even this is not all that was wrong with this decision.
Article 15 of Uber’s Service Agreement with its drivers houses both a choice of law and arbitration agreement. In it, parties agree that their agreement shall be exclusively governed by and construed in accordance with the laws of The Netherlands and that Any dispute, conflict or controversy…including relating to its validity, its construction or its enforceability shall use mediation and arbitration using ICC Rules seated in Amsterdam. Dutch law applies to this arbitration agreement, either because of express designation, or because the parties chose Amsterdam as the seat.
Justice Nordheimer accepted this clause as a choice of law clause and an adjudication clause. He even acknowledged that the arbitration agreement establishes “a foreign law that will be applied in that adjudication”. But he still applied Ontario law.
Under Canadian law, courts treat foreign law as questions of fact. And in the lower court, Justice Perell seems to have admitted as fact that Dutch law governed the Service Agreements between Uber Drivers and Uber. While the rule that appellate courts must accept facts as found by the lower court does have an exception, Justice Nordheimer did not avail himself of this exception and even if he had, he would have had to confront the content of Dutch law. But he didn’t. Consequently, Justice Nordheimer was tied to the lower court’s finding that Dutch law applied.
Therefore, even accepting Justice Nordheimer’s view that the Court should first address an arbitration agreement’s validity, his analysis should have applied Dutch law. This holds even under the domestic arbitration act.
Justice Nordheimer did not deal with whether choosing Dutch law was also unconscionable, focusing instead entirely on the dispute resolution mechanism. Once again, Dell comes into play. Because Dutch law should have applied, the question was one of fact about how Dutch law treats unconscionability – a question that a court cannot address superficially. Hence, Justice Nordheimer should have referred this case to the arbitral tribunal.
A better test to enforce arbitration agreements
Finally, it is dangerous to draw a straight line between invalidity, under s. 7(2) of Ontario’s Arbitration Act, and unconscionability, when dealing with clauses relating to future, knowable but specifically unknown disputes. Exclusion clauses are similar to arbitration agreements as they also seek to address breaches of future knowable, but specifically unknown facts. If Canadian courts are searching for an appropriate test to apply when seeking to enforce (or not) arbitration agreements, they should look to the Supreme Court’s decision in Tercon, where Justice Binnie offered a very helpful three part test, which includes an unconscionability component at stage two. Importantly, under this test, unconscionability is assessed at time of contracting, which eliminates outcome bias when viewing a transaction’s fairness.
In his decision, Justice Nordheimer relied on Titus v. William F. Cooke Enterprises Inc. (which itself relied on Cain v. Clarica Life Insurance Co.) to analyze whether the arbitration agreement between Uber and Mr. Heller was unconscionable. Those cases dealt with known disputes at the time of entering into release agreements. An important reason why Justice Nordheimer found the clause in Uber was unconscionable related to the low amount in dispute compared to the higher cost to commence the arbitration. This is classic outcome bias. Adopting Justice Binnie’s three part test to arbitration agreements is more appropriate as it permits what invalidity under s. 7(2) seeks to address: 1. Does the agreement apply to the dispute? 2. If so, at the time of signing, was it unconscionable (or another invalidity issue) to include this arbitration agreement? 3. If not, do any policy reasons exist to justify refusing to enforce the arbitration agreement?
Applying this test would allow a decision maker to consider whether the more expensive arbitration option was fair at the time of agreement, in light of all circumstances including other options (like Skip the Dishes), regardless of the cost to commence arbitration at the time the dispute arises.
The correctness of Justice Nordheimer’s decision is chimerical because he omitted many fundamental steps. This makes identifying his errors more challenging because it requires knowing how arbitration is intended to interact with courts, which is an area Canadian courts are still exploring. Nevertheless, the ONCA decision misapplied the law and the SCC precedent by not sending the matter to the arbitral tribunal.
The Hualon decision
I picked this case for two reasons. First, it reflects “best practices” on how courts should interpret uniform international arbitration legal instruments and second, a Canadian penned the decision thus proving that arbitration is not beyond Canadian judges.
Like Ontario, BVI is a Model Law jurisdiction and has also adopted the New York Convention. In the BVI Court, the Claimant (Hualon), applied to stay its claim pending the outcome of an international arbitration commenced against the defendant, Marty Limited—an application as routine as they come in arbitration, but containing an interesting twist.
Ordinarily, the party commencing a court action is the same party attempting to bypass the arbitration agreement. But in Hualon, the very same party who commenced the court action (Hualon) was now attempting to stay the court action it commenced. The Respondent (Marty) opposed the stay, claiming, in particular, that by commencing the court action, Hualon repudiated its agreement to arbitrate.
Justice Leon accepted Hualon’s position that it only discovered the arbitration agreement after commencing the court action. Thus, he found that the court proceedings and subsequent referral application were not tactical manoeuvres from a party seeking some untoward advantage.
As a result, Justice Leon granted the application to stay the court proceedings and referred the parties to arbitration, in particular to allow the arbitral tribunal to decide whether it had jurisdiction.
While Hualon deals with very different facts, the decision is a study on how to apply uniform legislation like the Model Law and the New York Convention. For purposes of this short comment, I will limit myself to two important lessons Canadian judges can take from this decision.
Respect for Competence-Competence
Justice Leon respected the competence-competence doctrine. This doctrine holds that when parties raise questions related to an arbitrator’s competence (or jurisdiction) the arbitrators may decide these questions, and should decide them in the first instance. This is necessary for arbitration to function effectively.
If competence-competence were not in place, parties could shirk their promise to arbitrate by raising jurisdictional arguments and challenging the tribunal’s jurisdiction. Competence-competence addresses this possible abuse. If a party really wishes to attack an arbitral tribunal’s competence or jurisdiction, it must address this concern to the arbitrators themselves. This way, a jurisdictional attack cannot become a ploy to avoid arbitration. As explained above in the Uber discussion, this doctrine is embedded in Canadian arbitration law, including domestic and international arbitration legislation.
Justice Leon further upheld competence-competence by finding that not only should arbitrators decide their own competence – they should be first in line to do so. He referred the parties to the arbitral tribunal, thereby allowing the tribunal to decide if it had jurisdiction.
Hualon supports arbitration as it respects the parties’ choice to first use arbitration to resolve disputes, even jurisdictional disputes. It also respects the arbitral tribunal’s role. When a court first resolves questions that properly should go to the arbitral tribunal, courts subvert the arbitral process.
Unconscionability calls into question a party’s consent. And at its highest, Mr. Heller’s (the Uber driver) argument was that he had not consented to arbitrate disputes with Uber via mediation and arbitration in Amsterdam. Alone, this would have made Nordheimer JA’s reasoning a bit easier to understand. Except, Mr. Heller was not arguing that he did not know that he had entered into a contract containing an arbitration agreement (conceding that, prima facie, an agreement to arbitrate exists). He argued that Uber’s terms contradicted Ontario’s EAA and this violation rendered the arbitration agreement invalid.
Had Justice Nordheimer correctly applied competence-competence, an arbitral tribunal would have first decided on the agreement’s validity. After all, both Ontario’s domestic and international arbitration legislation allow arbitrators to resolve validity questions.
Looking to other jurisdictions to inform his decision, per Model Law 2A
The Model Law is a legal instrument intended to apply uniformly throughout all jurisdictions. Consequently, under the Model Law, decision-makers are expected to look to other jurisdictions to see how they interpret same or similar provisions. The Model Law’s 2006 amendments, adopted in Ontario since 2017, sought to put this point beyond doubt by adding article 2A to the Model Law.
In his decision, Justice Leon referred to the travaux and cited cases from, inter alia, BVI, Canada, England & Wales, Hong Kong, and Singapore. He even cited to authorities from around the world. In short, he did exactly what the Model Law hopes decision-makers will do: respect the Model Law’s international origins and its need to promote uniformity in application. Citing to jurisdictions that have reached similar outcomes respects the purpose behind an international arbitration system.
Ontario’s current international arbitration legislation, the ICCA, which came into force in 2017, includes article 2A. The ICCA applied in Disney and arguably should have applied in Uber. And if the ONCA had applied the ICCA correctly, it would have relied on article 2A, which would require the Court to look beyond domestic law. In Disney this may have opened the Court’s eyes to seat theory, allowing it to understand the interaction between Ontario’s mandatory arbitration provisions and the parties’ choice to incorporate the English Arbitration Act. For Uber, it could have directed Justice Nordheimer to better understand competence-competence, allowing him to recognize that Dutch law should have applied to the arbitration agreement and to his “unconscionability” analysis.
When parties enter cross-border commercial relationships, private arbitration is often chosen as the method to resolve disputes.
Arbitration’s integration into the Canadian legal framework is critical for Canada’s economy. Courts play an important role in this growth, including through their supervisory role over arbitration.
At the time of writing, Uber has successfully sought leave to appeal the ONCA’s decision to the Supreme Court. But it shouldn’t take our highest court to keep arbitration on course, as it was recently forced to do in Telus Communications Inc. v. Wellman. Instead, our courts should make better efforts to learn this critical area. And they have many places to look.
* Anthony Daimsis is a professor at the University of Ottawa, Faculty of Common Law, Director of its National Program and its Mooting Program, and is an Associate Door Tenant at Littleton Chambers in London, England.
 The Walt Disney Company v. American International Reinsurance Company, Ltd., 2018 ONSC 1108 (“SC-Disney”), and 2018 ONCA 948 (“ONCA-Disney”).
 2019 ONCA 1.
 Claim No. BVI HC (COM) 2014/0090.
 For example, it is practically impossible to miss the seminal seat theory case from England: Union of India v McDonnell Douglas Corp  2 Lloyd’s Rep 48. Virtually all arbitration books discuss competence-competence, seat theory, and uniformity. For example, Gary Born’s International Commercial Arbitration: Commentary and Materials, Nigel Blackaby and Constantine Partasides, Redfern & Hunter on International Arbitration, Fouchard, Gaillard, Goldman on International Commercial Arbitration, Brian Casey’s Arbitration Law of Canada: Practice and Procedure (now in its third edition), to name only a few.
 Arbitration Act 1996, 1996 c. 23.
 ONCA-Disney at para. 2.
 S.O. 2017, Chapter 2 Schedule 5. The ICCA adopts the United Nations Commission on International Trade Law (UNCITRAL), UNCITRAL Model Law on International Commercial Arbitration, Annex I, UN Doc. A/40/17 (1985) and amended in 2006 (“Model Law”) and the Convention on the Recognition and Enforcement of Foreign Arbitral Awards, New York, 10 June 1958, 330 U.N.T.S. 38 (“New York Convention”).
 The Analytical Commentary makes this point, see Doc. A/CN.9/264, pp. 44-45.
 SC-Disney at para. 15. Although Justice Pollack placed party autonomy at the forefront, in reading the decision, it is not entirely clear if the parties did, in fact, agree on this. ONCA seemed equally suspect (see ONCA-DISNEY at para. 3).
 SC-Disney at para. 24.
 The ICAA internalizes “seat theory” at Model Law Article 1(2).
 ONCA-Disney at para. 4.
 The Constitution Act, 1867 (UK), 30 & 31 Victoria, c 3.
 S.O. 1991, c. 17
 Uber at para. 21.
 2A(1) reads: In the interpretation of this Law, regard is to be had to its international origin and to the need to promote uniformity in its application and the observance of good faith.
 Applying Ontario’s international arbitration act (and in particular Model Law article 2(A)) Justice Nordheimer would have noted that other jurisdictions respect governing law provisions, like the one found Uber’s Service Agreement.
 Uber at para 39.
 Dell Computer Corp. v. Union des consommateurs,  2 S.C.R. 801, 2007 SCC 34 (“Dell”) at para. 84.
 My use of the word “could” is deliberate since a tribunal could also assess questions of law in a shared model like the one the Model Law endorses. And while a court could also address questions of fact and mixed fact and law, this assumes the parties have agreed to this.
 Dell at para. 85. Similarly, in the seminal Mitsubishi v Soler case, the US Supreme Court rejected the innate hostility towards arbitrators by recognizing arbitrators are equally competent as judges to resolve many disputes including, in that case, anti-trust claims. See Mitsubishi Motors v. Soler Chrysler-Plymouth, 473 U.S. 614 (1985), at paras. 14-19.
 Gary Born, International Commercial Arbitration, 2nd edition, Kluwer Law International, New York 2014, pp 1096-1097.
 Although Canada does not follow the US model of neatly distinguishing between procedural and substantive unconscionability, one could argue that statute-based and court-based rules fit within substantive unconscionability, while factual unconscionability is closer to procedural unconscionability.
 A point Justice Perell in the lower court decision acknowledged. See Heller v. Uber Technologies Inc., 2018 ONSC 718, at para. 47 (“SC-Uber”).
 Unlike arbitration –and in particular under the ICC Mediation Rules—mediation is voluntary and hence a court does not compel or “refer” parties to mediation, absent a court rule. We must assume the parties in Uber voluntarily chose to forgo mediation. Alternatively, until a decision on the enforceability of Article 15 is reached, any obligation to first use mediation has not crystalized.
 Under article 18(2) of the ICC Arbitration Rules, the arbitral tribunal could have conducted the hearing anywhere, including Toronto. Another fact the Ontario courts seemed to overlook.
 Uber at para. 63.
 Hunt v T&N plc,  4SCR 289 at 308. See also The Evidence Act, SS 2006, c E-11.2, s 4(2).
 SC-Uber at para. 21.
 Stephen Pitel and Nicholas Rafferty, Conflict of Laws, 2ed (Toronto: Irwin Law, 2016) at page 251.
 Tercon Contractors Ltd. v. British Columbia (Transportation and Highways), 2010 SCC 4,  1 S.C.R. 69, at 127-141.
 2007 ONCA 573 (CanLII), 284 D.L.R. (4th) 734, at para. 38.
 (2005), 384 A.R. 11 (CA); 367 W.A.C. 11.
 Uber at para. 68.
 The relevant provisions are: article 8(1) of the Model Law and article II(3) of the New York Convention.
Model Law article 8(1):
A court before which an action is brought in a matter which is the subject of an arbitration agreement shall, if a party so requests not later than when submitting his first statement on the substance of the dispute, refer the parties to arbitration unless it finds that the agreement is null and void, inoperative or incapable of being performed.
New York Convention article II(3):
The court of a Contracting State, when seized of an action in a matter in respect of which the parties have made an agreement within the meaning of this article, shall, at the request of one of the parties, refer the parties to arbitration, unless it finds that the said agreement is null and void, inoperative or incapable of being performed.
 2019 SCC 19.