Bolch Judicial Institute
Duke Law School
In recent years, governments from the state of Delaware to the Emirate of Dubai have created institutions specially designed to adjudicate transnational commercial disputes. These institutions are hybrids between courts and arbitration, or “arbitral courts.”1 Arbitral courts seek to adapt the most popular features of arbitral tribunals and courts in order to accommodate the growing challenges of such cases.
Arbitral courts mimic arbitration’s traditional features to some degree. They have internationally well-regarded judges who may also work as arbitrators. They claim the neutrality, the expertise, and sometimes the privacy and confidentiality of international arbitration. Unlike arbitration, however, they bind third parties, develop transnational law, and wield the power of the state.
There is much to applaud about the innovation of arbitral courts. But it is important to consider what limits should cabin this innovation. Arbitral courts unsettle the traditional distinctions between public and private adjudication, and this blurring has significant consequences not only for understanding the state of the evolving international judicial system, of which U.S. courts have historically been an important part, but also for the future of legitimacy and transparency in dispute resolution around the world.
Arbitral courts often claim legitimacy on grounds that combine arbitrators’ and courts’ claims to legitimacy. The legitimacy of arbitration mostly flows from parties’ consent to the arrangement, whereas courts’ legitimacy, at least those courts situated within democracies, derives more broadly from social compacts and customs, including from the democratic legitimacy of the state.
Standing at the crossroads of public and private adjudication, however, arbitral courts could abuse their position by exercising jurisdiction in contexts beyond the scope of what gives them legitimacy (e.g., parties’ consent) or by closing off public access. There is a significant risk that they will do both of these things. Like arbitration centers’ rules, arbitral courts’ rules tend to be flexible. The courts often have substantial discretion over issues like whether to join third parties who have not consented to jurisdiction and whether to grant parties’ requests to keep the proceedings and decisions confidential, which tends to result in keeping proceedings secret. Exercising exorbitant jurisdiction and proceeding in secret, however, could undermine an arbitral court’s reputation for evenhandedness, its perceived legitimacy, and its potential to develop transnational law.
In light of these risks, I offer two suggestions for arbitral courts looking to build their own legitimacy and to contribute to improvements in judicial institutional design. First, arbitral courts should restrict their jurisdiction in light of their hybridized source of legitimacy that draws on their resemblance to both a court and an arbitral tribunal. Second, arbitral courts should prioritize the public nature of proceedings and decisions and not defer to parties’ requests for confidentiality.
Arbitral courts are at the vanguard of international commercial dispute resolution. They are domestic institutions designed to hear cases involving actors or controversies that cross borders. As such, they are an important addition to the “international judicial system.”2 They represent an important trend in transnational and commercial litigation that should be watched carefully, and that hold potential lessons in institutional design relevant to courts everywhere.
Arbitral courts are domestic courts that have the following arbitration-like characteristics: they (1) allow party autonomy over procedures, (2) proceed in English, (3) permit parties to opt into confidentiality, and (4) exercise jurisdiction based on consent, often without further connections to the locality. They often (5) employ foreign judges, (6) have judges sit in three-judge panels, (7) offer opportunities for foreign lawyers to appear without local counsel, (8) allow parties to opt out of the right to appeal, and (9) are willing to enforce parties’ selection of non-state law to govern their dispute. Some of these traits are common to all arbitral courts, and some are typical but not necessarily found in each example of an arbitral court. Regardless of the precise collection of arbitration-like characteristics, arbitral courts reveal a trend that extends beyond recognized ways in which courts, even commercial courts, have been catering to private parties and their disputes, mimicking arbitration, and, as we shall see, neglecting their roles as public institutions.
Using the term arbitral courts recognizes that these entities are domestic courts. They are not international courts like, for example, the International Court of Justice. Individual states3 create and fund them (at least as a formal matter). Nor are they actually private arbitral tribunals. They render binding decisions enforceable by the power of the state. They can issue subpoenas and interim relief. They can join or bind non-consenting third parties. Under common law traditions, they can declare law and establish precedent.
But arbitral courts nevertheless closely resemble arbitration.4 Arbitral courts build on the efforts of commercial courts to respond to parties’ preferences for speed, flexibility, and expertise.5In a number of ways, procedure in commercial courts has become increasingly privatized — through managerial judging, court-annexed arbitration, and increased party control over procedures. Arbitral courts take these efforts several steps further. Arbitral courts may employ foreign judges, allow foreign lawyers to appear before them, and permit parties to opt out of appeals. They operate in English (even in non-English-speaking countries). They allow parties to choose which forum hears the dispute (regardless of the forum state’s connection to the dispute), which procedures apply, whether the proceedings or the resulting decision will be kept confidential, and what law governs the dispute, potentially even if parties select non-state-created law, like general equitable principles, or rules articulated by organizations like the United Nations Commission on International Trade Law (UNCITRAL).
Delaware provided an early and important example of an arbitral court. In 2009, Delaware enacted a statute allowing Chancery Court judges to act as arbitrators.6 For controversies involving at least one Delaware business entity, no consumers, and amounts in dispute over $1 million, the parties could agree to have a Chancery Court judge arbitrate their dispute. The proceedings would be confidential and held in the Delaware courthouse for a filing fee of $12,000, plus $6,000 for each additional hearing day. Regular Chancery Court procedure and evidence rules would apply, but the parties could agree to modify them.7 The judges could grant any remedy they “deem[ed] just and equitable and within the scope of any applicable agreement of the parties.”8 The losing party could appeal the “order of the Court of Chancery” to the Delaware Supreme Court, but subject to Federal Arbitration Act standards of review. The arbitration petitions and decisions would be confidential, but once appealed they could become part of the public record.9 Delaware designed the statute, Chief Justice Myron Steele explained, “to keep the United States, and in particular, Delaware, competitive in international business dispute resolution.”10
In 2013, a panel of the Third Circuit declared that these “government-sponsored arbitrations” violated the First Amendment’s right of public access to trials because of their confidential nature. The Third Circuit judges in Delaware Coalition for Open Government v. Strine debated whether the Delaware statute created a court that had some arbitration-like features (like confidentiality, optional procedural rules, limited appellate review), which would require public access, or an arbitral tribunal that had some court-like features (Delaware Chancery judges, Delaware courthouse), which would not. In fractured decisions, two of the three judges thought Delaware had unconstitutionally created confidential courts.11 Delaware’s courts had to be open to the public.12
The Third Circuit thus thwarted Delaware’s attempt to create a court-arbitration hybrid — what this article calls an “arbitral court.”13
The Third Circuit’s decision reflects both conventional civil procedure theory and arbitration theory about the dividing line between courts and arbitration. Courts are public, “procedurally rigorous,” and state-sponsored; arbitration is private, “faster and cheaper but with fewer procedural safeguards.”14 Courts’ authority derives from the state; their power extends as far as the state’s. Arbitration, by contrast, is a “creature of contract.”15 The parties’ agreement both defines and limits arbitral tribunals’ authority. While scholars have recognized a convergence of procedures in different fora16 and bemoaned both the privatization of court procedure17 and the judicialization of arbitration,18 the understanding was that courts and arbitration stay in their lanes.19
While the Third Circuit saw Delaware’s arbitral court as a bridge too far, the idea has had a more positive reception elsewhere. International commercial courts that mimic traits from arbitration have been established in Dubai (2004), Singapore (2015), and the Netherlands (2019).20 In the same period, tax havens such as Bermuda, the British Virgin Islands, and the Cayman Islands, the place of incorporation for many foreign firms, have established new business courts that “resemble commercial arbitration.”21 This article considers Delaware’s government-sponsored arbitration experiment, some international commercial courts,22 and these offshore courts to be “arbitral courts.”
Arbitral courts shift and blur traditional boundaries between public and private adjudication. The lessons from arbitral courts are, in part, positive lessons about the power of procedural innovation and forum shopping, which I have explored in previous work.23 But as institutions at the crossroads of public and private adjudication, arbitral courts reveal not only the possibilities but also the limits of experimentation and party autonomy over procedure, especially over questions of confidentiality.
By combining attributes of both litigation and arbitration, as Hiro Aragaki has explained, courts like the arbitral court in Singapore reject “an either/or choice between public and private adjudication; instead, they think of dispute resolution holistically, all the while borrowing one device from one process and glomming it on to another without so much as an afterthought.”24 Courts and their designers seem to be throwing traditional distinctions to the wind.
But can arbitral courts coherently reject public/private distinctions in all senses? While there is a flexibility in the distinction between public and private adjudication in many respects, there are and should be limits. Here I focus on two.
First, there should be consistency between a court’s claim to legitimacy and its jurisdictional reach. If a court claims legitimacy based on the consent of the parties before it, then its jurisdiction should be so limited, just as an arbitral tribunal’s jurisdiction would be. That consent should not justify the court reaching beyond what would otherwise be the limits of its jurisdiction as it applies to parties and disputes beyond what the consenting parties have agreed to.
Second, arbitral courts, like all courts, are public institutions. When courts are not transparent about their proceedings or decision-making, or when courts entrust parties with questions of confidentiality, they cross a line that compromises their effectiveness — in dispute resolution, in law making, and beyond. Dispute resolution behind a veil of opacity intentionally hides these functions from public observation and, thus, dispenses with any obligation of justification.25Trust in arbitral courts then becomes solely a function of blind trust. Transparency, then, contributes to the growth of arbitral courts by promoting their legitimacy, effectiveness, and their ability to develop transnational law.
To attract cases and to ensure their decisions are widely enforceable, arbitral courts will need to establish both legal and sociological legitimacy.26 Legitimacy is a complex concept.27 To function, a court needs the public to perceive it as both “playing an appropriate role in . . . governance” and “making its decisions based on ‘law,’ not ‘politics’ or ‘personal preferences.’”28 Put another way, both courts and arbitration need both sociological legitimacy and legal legitimacy. The sociological legitimacy of a court depends on whether the public views the court “as worthy of respect and obedience.”29 Legal legitimacy is established by using accepted interpretive methods and fair procedures.30
The central source of international commercial arbitration’s sociological and legal legitimacy is freedom of contract. As discussed below, decision-makers (arbitrators) also lend sociological legitimacy because of who they are and how they are chosen. The wide respect for freedom of contract and the structure of arbitration has led to an international structure of support built by international treaties, national courts, and private interests.
Courts generally have different sources of sociological legitimacy — including from the state. In democracies, for example, courts have democratic legitimacy.31 In non-democratic states, courts sometimes lend legitimacy to the state as opposed to the other way around. That is, “to varying degrees, [authoritarian regimes] also attempt to make up for questionable procedural legitimacy by preserving judicial institutions that give the image, if not the full effect, of constraints on arbitrary rule.”32
Arbitral courts rely both on arbitration’s sources of legitimacy as well as on the authority of the state. That is, like arbitration, arbitral courts seem both sociologically and legally legitimate in large part because parties have chosen to have their disputes adjudicated there.33
For decades now, it has become commonplace that courts can adjudicate disputes based on forum selection clauses even if the parties and the dispute have no ties to the forum.34 Parties can consent to jurisdiction in courts just like they can in arbitration, thereby giving the court or arbitral tribunal jurisdiction over a set of parties and a set of disputes that they might not otherwise have.
When they do so, it would seem logical that arbitral courts, like arbitration, would limit their jurisdiction to the case that the parties chose to submit to them. But this is not usually the assumption. Ordinarily, once parties have consented to a court’s jurisdiction, the court exercises the full force of its powers as an arm of the state — including the power to issue subpoenas and injunctive relief and to consolidate cases not subject to the parties’ forum agreement. Moreover, unlike what typically happens in arbitration, the parties’ consent does not limit the court’s jurisdiction to the parties who have consented to that jurisdiction. That is, consent-based court jurisdiction results in cases where ordinary courts can issue subpoenas to third parties, join third parties, and otherwise consolidate cases — even when there is no other territorial basis, beyond the presence of the consenting parties, for the court’s jurisdiction. Indeed, part of courts’ attractiveness as a forum is that they can bind third parties and adjudicate some kinds of cases (like business torts) that might not be able to be subject to arbitration.35
This is true for arbitral courts as well, even if their only claim to jurisdiction is the original parties’ consent. For example, the Singapore International Commercial Court (SICC), established in 2015, is one of the most celebrated arbitral courts. Once the SICC has jurisdiction over a case because two contracting parties consented to have the SICC hear their contractual disputes, the court has the full power of the state to include in the proceedings additional parties who may not have consented.36 Indeed, the SICC rules grant the court expansive authority to join non-consenting parties, including the authority to name them as additional plaintiffs or defendants — even if, apart from this joinder, they have no other connection to Singapore and Singapore would otherwise lack judicial jurisdiction over them.37 Such joinder is not typically available in arbitration; the arbitral tribunal’s jurisdiction is set — and limited — by the scope of the parties’ agreement.
This result in arbitral courts like Singapore’s is, at best, awkward (in all cases), and at worst, illegitimate (in cases where the arbitral court has no other basis for exercising jurisdiction over the non-consenting third party). It is unclear, however, how often this extension of jurisdiction happens38 or whether the public would know if it did occur, in light of arbitral courts’ sometimes non-public dockets.
The point, however, is that questions of jurisdiction over third parties test the distinctions between the legitimacy of the court acting as a public dispute resolution forum on the one hand or a private dispute resolution forum on the other. Arbitral courts suggest that the public/private boundary may be shifting, in international commercial disputes, to one defined by the line between authority based on parties’ consent and authority based on state sovereignty. Arbitral courts are trying both to straddle that line and to circumvent it. State sovereignty alone, however, does not establish judicial jurisdiction over everyone everywhere. Nor does it establish judicial legitimacy, especially in states that seek to draw legitimacy from the strength of their judicial institutions.39 It is circular to ask the courts’ consent-based legitimacy to support the state’s legitimacy as a sovereign, while also asking the state’s power to extend the arbitral court’s power beyond its original source: the parties’ consent.
This article, therefore, recommends that arbitral courts limit the scope of their jurisdiction — particularly with respect to questions of expansive jurisdiction, like jurisdiction over third parties — based on the scope of contract granting them jurisdiction.
The same limits — and a similar recommendation — echo in the area of enforcement. One of the most often cited reasons for choosing arbitration over litigation is the easy availability of enforcement of arbitral awards around the world. But this distinction may be eroding. Based on a trio of treaties,40 the difference between easy enforceability (traditionally associated with arbitral awards) and more demanding scrutiny (traditionally associated with court judgments) may ultimately depend on whether the parties have agreed to the forum’s jurisdiction — not whether that forum was an arbitral tribunal or a court. Under this framework, it will be consent, or its absence, that distinguishes between ready international legitimacy and suspicion — not the difference between an arbitral tribunal and a court. If that becomes the norm over time, that may further weaken distinctions between arbitration and litigation in transnational disputes and governance more generally.
But as is the case with jurisdiction, the scope of consent should limit the scope of enforceability. For example, if the SICC were to use its authority to join non-consenting third parties, over whom the court otherwise would not have jurisdiction, then foreign courts should not enforce the resulting judgment against the third party.41
In arbitration, the decision-makers (arbitrators) also lend sociological legitimacy because of who they are and how they are chosen. Arbitrators appear more neutral because they are not state actors and possibly do not share a nationality with one of the parties, in contrast to judges on traditional courts. Many arbitrators are well-regarded experts in their fields. Moreover, they are chosen by the parties, and therefore even the losing party has helped constitute the tribunal and may feel like it had an advocate during the decision-making process. The chair of an arbitration panel — the third arbitrator chosen by the parties’ chosen co-arbitrators42 — also has legitimacy based on perceived neutrality because of this selection process.
Arbitral courts rely not only on consent as an arbitration-like basis for legitimacy, but also on the personal legitimacy of the decision-makers to signal the courts’ independence. Thus, Singapore, Dubai, and the Caymans have hired international judges from the UK and elsewhere — to bring with them the credentials, trustworthiness, and legitimacy that Delaware sought to sell with its judges and their excellent reputations.43 As non-nationals, these foreign decision-makers may quell potential concerns that these courts will exhibit bias in favor of local parties or the local government. They may thus enhance the perceived legitimacy and independence of the new court.44 Relying on the legitimacy of foreign experts as decision-makers is once again appropriated from international arbitration.
Arbitral courts’ staffing also tries to address common challenges to arbitration’s legitimacy. For example, arbitral courts adopt their own ethical rules, warding off arbitration’s problem of lacking an applicable ethics code.45 And unlike arbitral tribunals, which commonly allow parties to have a role in choosing their arbitrators, arbitral courts assign judges to particular cases. This practice seems aimed at increasing the likelihood of independence and objectivity and preventing objections that the arbitral court may be beholden to the parties before it.
Whether this appropriated legitimacy will ultimately be judged as credible will depend on how cases play out in these courts — something we can observe only if arbitral courts commit to transparency. Transparency is likewise key to knowing whether arbitral courts will operate independently. If government, government-connected, or otherwise high-influence parties come before arbitral courts, will the courts maintain their neutrality? As the expression goes, “only time will tell.” But in order for time to tell, the record cannot be shrouded in secrecy.
These courts are still new and must be watched for evidence. The Dubai International Financial Centre (DIFC) courts, one of the oldest examples of this new wave of arbitral courts, have had mixed results. As Matthew Erie has documented, DIFC Courts have ruled in favor of the government bodies that have appeared before them, but they have also ruled against quasi-government corporations.46 This example yields hope but does not totally alleviate skepticism — and there must still be transparency in order to monitor arbitral court independence.<
The success of arbitral courts’ hybrid approach to legitimacy thus will depend in large part on how much of the arbitral courts’ operations the public can see. This is true not only for the publication of opinions, but also for the transparency of process, dockets, access, and other dealings. As the Delaware experiment revealed, keeping confidential the working of arbitral courts tends to undermine the institutions’ legitimacy. Questions of publicity and confidentiality are therefore of utmost importance, but they may face resistance from another signature feature of arbitral courts: party autonomy.
Arbitral courts purport to be public institutions. One common characteristic of public courts is that they have a consistent set of procedures and rules that apply to all who come to them. Nevertheless, arbitral courts give parties considerable choice and control over procedural and evidentiary rules. In many instances, the parties’ interests will be antagonistic towards one another in a way that balances to yield normatively acceptable procedural rules. For example, the plaintiff might want extensive discovery, the defendant might want minimal discovery, and in contracting for procedure, they might reach a compromise solution.
On confidentiality decisions, however, experience teaches that the parties’ interests will likely be aligned in favor of confidentiality. But public access — to courts’ proceedings, records, and decisions — would further the long-term institutional interest of the forum and the law. In regular arbitration, parties are free to agree to keep their disputes — including the proceedings and resulting decisions — confidential and private. Confidentiality is not an inherent attribute of international commercial arbitration, but it is an available option.47 And, when offered, parties often choose to keep their disputes secret. That is generally appropriate because the parties have chosen a private way of settling their disputes and that resolution applies only to them. Courts, by contrast, ordinarily limit opportunities for confidentiality of regular proceedings and of judicial decisions, and courts and scholars alike urge the importance of “open justice.” This openness is important for a number of reasons, including that the courts make law not just for the litigating parties but also for others, they exercise government power, and they are publicly funded and provide a public good that the public should be allowed to monitor to prevent corruption and misuse.
Permitting party control over decisions about confidentiality can make arbitral courts much closer to private, rather than public, institutions. The pitfalls of trusting the openness of arbitral courts to the parties have been demonstrated in the court context, such as in the history of the opioid litigation,48 as well as in the international commercial arbitration context, where calls for more institutional transparency have run up against party preferences.49 Control over confidentiality choices is a little like control over forum choices — the allocation of decision-making authority typically decides the outcome.50
The traditional distinction between “private arbitration” and “public courts” has long been eroding, particularly as courts offer parties more autonomy regarding choice of forum, choice of law, and discovery. But arbitral courts upend the distinction completely. Judges, who usually control decisions about confidentiality powered by strong presumptions against it, may still be the ultimate decision-makers. But arbitral courts may yield a trend toward party control over confidentiality decisions with minimal judicial supervision, powered by presumptions favoring the parties’ preferences. Those preferences are likely to favor confidentiality.
Structurally, arbitral courts may be likely to favor pleasing the parties and accommodating those preferences. Arbitral courts have put themselves in this position to cater to their customers — i.e., potential parties to international disputes. Offering confidentiality is a form of “forum selling”51 as a way to compete with arbitration to attract parties to select the arbitral court in their forum selection clause or otherwise choose the arbitral court for disputes.
The originators of the term “forum selling” suggested that the practice was problematic when courts were selling themselves to plaintiffs who maintained unilateral control over forum choices, but not when parties mutually agreed on a forum in a forum selection clause.52 In that latter situation, the authors assumed parties would choose the best courts, and courts would be driven to provide quality proceedings that would satisfy both sides.
But there are times when parties’ interests conflict with those of the court and the public. Joshua Karton has demonstrated that transparency presents such a conflict in international commercial arbitration.53 The institution needs transparency to sustain sociological and legal legitimacy in the long run, but it also needs to offer confidentiality to attract parties in the short term. In arbitral courts, these competing needs exist to an arguably even greater extent. Arbitral courts are new, and therefore need participants. But they also need to publicize what they are doing to establish themselves. In democracies, such open justice is necessary to bolster judicial independence, build legitimacy, and allow for public oversight. Arbitral courts in non-democratic states have an even higher burden to demonstrate to the public that they are independent and follow the law. And ultimately, the world of commercial parties who might choose to litigate their disputes there will also be attracted by the legitimacy offered by transparency, and not just the convenience provided by secrecy.
One might wonder whether arbitral courts’ rules about confidentiality are or should be constrained by some higher order law. While the UK Supreme Court has recently reaffirmed the principle of “open justice,” it is not clear whether and how this rule will bind or persuade arbitral courts in Singapore, Dubai, or Kazakhstan, even if those courts operate under the common law tradition. The Third Circuit held that the Constitution enshrines a similar principle, thwarting Delaware’s attempt to create a confidential arbitral court. But it probably would have been possible for Delaware (or another interested state) to circumvent such a ruling. A revised statute, for example, could make confidentiality available at the judge’s discretion, and then set (or allow judges to set) a low threshold for granting confidentiality requests (perhaps even not requiring the request to be bilateral). Such a setup might have satisfied the Third Circuit. It is also possible that another set of federal judges might agree with the dissent rather than the majority in Delaware Coalition.
Regardless of the answer to the constitutional question, the Third Circuit was right in its more basic conclusion that confidentiality was the central problem with the Delaware arbitral court from an institutional design perspective. Its decision is also supported by both the judiciary’s interest in the public interest of access and by arbitral courts’ self-interest in supporting their budding reputations, legitimacy, and power.
Arbitral courts are widely celebrated as adopting the “best” of both worlds of binding adjudication: arbitration and litigation. But arbitral courts bring perils as well as promise. Arbitral courts that exercise jurisdiction beyond the scope of the parties’ consent could compromise their legitimacy and the likelihood that their judgments will be recognized and enforced by other courts. Arbitral court judges have the opportunity to be transparent and vocal in their neutrality — but will need to reject at least some requests for confidentiality to build their own reputations as well as the reputations of their institutions. Arbitral courts that proceed behind the dark veil of confidentiality, often requested by parties, could threaten their own legitimacy and eventually lead to their untimely demise. Liberal granting of parties’ requests for confidentiality will compromise not only the decisions made in particular cases but also arbitral courts’ abilities to develop law and shape global governance. Arbitral courts present a unique potential to develop transnational commercial law in a way that the confidentiality and lack of authority of traditional arbitration have thus far inhibited. If arbitral courts can contain the scope of their power within legitimate limits and if they are able to resist parties’ preferences for confidentiality, they may be able to achieve their promise.