Consolidation in Arbitration

While these principles are clear in theory, they are difficult to apply in practice. With regard to the standard for proving consent, for example, Stolt-Nielsen v. Animal Feeds International, 559 U.S. 662 (2010), held (5-4) that an agreement to arbitrate could not alone be the basis for finding intent to permit class arbitration. In that case, however, the parties stipulated that they had reached no agreement on the use of class actions, so it remains to be seen what proof the courts will require to establish consent where the parties dispute the issue. To what extent, for example, will language in an agreement requiring the application of a particular set of rules be sufficient to establish consent, or the lack of consent, to consolidation? Consent may reliably be inferred where an agreement explicitly adopts a set of rules that in turn explicitly allow consolidation. But uncertainties arise if arbitrators allow the use of consolidation in a manner that goes beyond the terms of a rule or on the basis of general objectives in a set of rules, such as efficiency or justice. Is it correct to infer no consent when the rules adopted by the parties have no consolidation provision?

Additional uncertainty results where agreements to arbitrate contain provisions that imply consent or lack of consent to consolidation. For example, discovery limitations (e.g., on the number of depositions), limits on the number or types of witnesses, a very limited number of hearing days, or the presence of confidentiality requirements may suggest that the parties contemplated individual hearings, especially when the consolidation sought is of scores or hundreds of claims.

Similarly, although the court has held in American Express v. Italian Colors Restaurant, 570 U.S. __, 133 S. Ct. 2304 (2013), and AT&T Mobility v. Concepcion, 563 U.S. 321 (2011), that provisions in arbitration agreements precluding the use of class actions must be enforced despite determinations that they are unconscionable under state law, or that individual claims would not be viable, such prohibitions may be unclear or less comprehensive than those present in the Amex and AT&T cases.

These uncertainties as to the standards for determining consent are greatly magnified by the fact that parties may choose to assign the decision on consolidation to arbitrators. Given the very limited judicial review of arbitrator decisions set out in Oxford Health Plans v. Sutter, 569 U.S. __, 133 S. Ct. 2064 (2013), arbitrators therefore can make decisions on consolidation that are inconsistent with the court's standards.

Furthermore, the court has yet to make clear whether the evidence needed to establish that the parties to an arbitration agreement assigned responsibility for deciding whether they consented to consolidation must meet the “clear and unmistakable” standard of First Options of Chicago v. Kaplan, 514 U.S. 938 (1995), rather than the generally applicable state-law principles that govern the formation of contracts. The court's distinction between substantive and procedural arbitrability in First Options, modified thereafter in Howsam v. Dean Witter Reynolds, 537 U.S. 79 (2002), led lower federal courts overwhelmingly to conclude that whether an arbitration agreement authorized class actions was a “substantive” issue rather than one that could be seen as “procedural.” But in Green Tree Financial v. Bazzle, 539 U.S. 444 (2003), four justices declined to apply the presumption in favor of judicial determination of arbitrability to the question whether an arbitration agreement prohibited the use of class actions. That issue, the justices explained, “concerns neither the validity of the arbitration clause nor its applicability to the underlying dispute between the parties. ' Rather the relevant question here is what kind of arbitration proceeding the parties agreed to.” Id. at 452.

The plurality opinion in Green Tree led arbitration services and arbitrators to conclude they could decide (without meaningful judicial review) whether parties had agreed to allow the use of class actions. Over 200 class actions were commenced, including some 100 under an AAA program. See, P. Christine Deruelle & Corey Berman, “The Future of Class Arbitration: Recent Supreme Court Jurisprudence on Arbitration Clause Interpretation and Enforceability,” Weil Class Action Monitor, p.9, n.9 (July 23, 2013).

Simultaneously, at least five federal circuit courts held that arbitrators could also decide whether the parties agreed to allow the use of consolidation (See, Robert W. DiUbaldo, “Evolving Issues in Reinsurance Disputes: the Power of Arbitrators,” 35 Fordham Urb. L.J. 83, 88 n.32 (2008)), leading to widespread efforts to consolidate cases or claims between 2003 and 2010. Most of the class action activity triggered by Bazzle ended after the Supreme Court decided in Stolt-Neilsen that class actions could not be imposed on parties with arbitration agreements silent on that issue, but efforts to impose (or find consent for) consolidation have continued, without authoritative resolution other than the very clear rule that a very clear delegation of the issue to arbitration is subject to very limited review even of very clearly wrong decisions.

Additional risks and uncertainty that arise when arbitrators insist on imposing consolidation include the possible effects of state laws based on the Revised Uniform Arbitration Act. The Act provides authority for courts to order consolidation on grounds similar to those that permit consolidation in litigation where the agreement to arbitrate does not prohibit consolidation. See, Section 10 (2000), reprinted in Stephen K. Huber & Maureen A. Weston, Arbitration: Cases and Materials, p. 491 (3d Ed. 2011) (LexisNexis). Do these statutes preclude arbitrators from consolidating claims, and if so for all purposes or only for the purpose of joint decision?

Another problem is reflected in the Sixth Circuit Court of Appeals' recent decision allowing a subcontractor to sue design professionals despite having participated in a consolidated arbitration that included the issues on which the subcontractor based its subsequent claim. See, W.J. O'Neil Co. v. Shepley, Bullfinch, Richardson & Abbott, No. 12-2320, 2014 U.S. App. LEXIS 16607 (6th Cir. Aug. 28, 2014). Will res judicata (i.e., claim preclusion) lose some of its effectiveness as a method for securing finality with regard to parties who are compelled to participate in consolidated arbitration? Several Justices have expressed doubt as to the propriety of requiring parties forced into class action arbitrations to opt out rather than requiring that they be included only if they opt in. So, even where consolidation may be properly mandated for a set of related claims, it may nonetheless be inapplicable to parties that have not consented to arbitrate those claims.

However well intended, therefore, efforts to mandate consolidation in arbitration without adequate proof of consent have been, and are likely to continue to be, little if any more productive than efforts to mandate nonconsensual class actions. Multi-party procedures are likely to prove useful in arbitration only through consent.

Advantages of Consent

What seems most important about all this, however, is not that consolidation has no future in arbitration. Rather, the principal lesson is that arbitrators and parties alike need to stop attempting to mandate full-fledged consolidation without proof of consent, and should instead exploit more systematically the potential benefits of using limited, pre-hearing consolidation based on actual consent. Consolidating arbitrations for all purposes, including hearing and decision, has a greater impact than even class actions in that they allow no right for individual parties to opt out. But unlike in class actions, consolidation can be used in arbitration for limited purposes. Arbitrations can be consolidated for general administration (thereby reducing the fees charged by arbitration services and enabling a single panel to administer the prehearing process); for document discovery (e.g., to compile a set of documents applicable to all the consolidated cases); for interrogatories (by making the evidence obtained usable in all proceedings in which it is admissible); for deposition discovery (e.g., to avoid multiple depositions of the same witness on the same issues); for some or all pre-hearing motions; and to permit the use by all parties in multiple proceedings of the same briefs and written witness statements. A similar, though mandatory, process is available in the federal courts through multi-district litigation designations. Finally, arbitrators can, in lieu of consolidating hearings, designate, hear and decide one or more typical claims rather than allowing the entire set of related matters to remain in limbo while the parties litigate whether all the relevant claims should be consolidated for all purposes.

Some of these processes may be undertaken without risk of serious challenge. A party opposed to consolidation for some purposes, moreover, may accept joint processing for more limited purposes to avoid unnecessary costs and duplication of effort. Deciding individual, representative claims could also significantly expedite the resolution of other, related claims. Parties in litigation faced with numerous, related claims in particular areas of commercial activity (such as construction, reinsurance, and franchising) sometimes designate specific cases for trial (or mock trial) in order to provide concrete indications of likely results, which they then use as guidance in settling or streamlining for trial the remaining claims. The same, rational process is likely to play out in arbitration; it is less dramatic than forcing joint hearings, but more effective at delivering concrete and useful results. This process is most likely to be effective in sets of claims that are each substantial enough to justify separate hearings. But it is also likely to be helpful in sets of claims that are each too insubstantial to warrant separate hearings, since joint processing and the decision of even one or a few such claims can lead to practical resolutions aimed at avoiding costs (including the possibility of attorneys' fees), the possible application of collateral estoppel, and the loss of commercial goodwill.

Finally, a danger of inconsistent rulings exists where different arbitrators are making decisions on the same issues. But few situations have been identified in discussions about consolidation in which parties have had to live, against their will, with inconsistent outcomes caused by inconsistent decisions of courts or arbitrators due to the unavailability of consolidation. The danger of inconsistent results is largely theoretical. Where inconsistent results actually occur, they may well reflect the fact that the correct outcome in those cases is itself uncertain, rather than any lack of competence or legal discipline on the part of the arbitrators or judges involved. Inconsistent results may well be no more than a signal to the parties of their need for realistic compromise.

Adopting a practical, consensus-based approach to consolidation in arbitration will not relieve arbitrators of their duty to decide legal issues. In some situations, parties may insist on arbitrator rulings on consolidation-related issues, and some may refuse to agree on any practical steps. In the absence of consent, however, arbitrators will be on firmer ground in requiring practical steps that do not prejudice the right to separate hearings and decisions than in issuing rulings that purport to impose the most consequential aspects of consolidation. The law is simply too uncertain to justify confidence in almost any one-sided determination. Even where the law is clear, such decisions are likely to lead to drawn-out, costly litigation during which all progress on the processing and resolution of the claims at issue is suspended. A flexible, consensual use of collective processing, rather than formal consolidation, is something arbitrators should seriously consider, regardless of their views on the legal merits of consolidation.

Finally, it is appropriate to note that the interest among parties and arbitrators in utilizing class actions and consolidation in arbitration reflects the continuing, potential utility of procedures adopted in the Federal Rules of Civil Procedure on the basis of some 150 years of litigation experience. The Federal Rules have real advantages in the sorts of litigation for which they are designed, whether conducted in court or in arbitration. They include provisions dealing with class actions and consolidation (though not with the issue of effects on parties that have not agreed to arbitrate). They are inapposite in contexts dissimilar to conventional federal litigation. But they are useful in the arbitration of disputes similar, if not identical, to those that arise in the federal courts. It should not be surprising that parties to such disputes that want to use arbitration for its many unique advantages (such as the choice of judges, applicable law, venue, confidentiality, etc.), may also want to utilize the Federal Rules where they appear advantageous compared to conventional arbitration procedures. (Federal Arbitration, of which the author is a founder and chairman, uses the Federal Rules of Civil Procedure as its default rules of arbitration, subject to appropriate modifications and to changes made by the parties.)

Abraham D. Sofaer is the George P. Shultz Senior Fellow at Stanford University's Hoover Institution and a founder and chairman of Federal Arbitration, Inc. He previously served as professor of law at Columbia University and as U.S. District Judge for the Southern District of New York. This article also appeared in the New York Law Journal, a sister publication to this newsletter.

Consolidation is one of several ongoing battlefields in the development of arbitration in America. Consolidating arbitrations among different parties can reduce costs, enhance efficiency and avoid inconsistent decisions. In practice, however, attempts to consolidate arbitrations without the consent of affected parties has resulted in costly litigation, long delays, and legal uncertainties that undermine the benefits of arbitration. Rules do exist concerning the proper use of consolidation, but they are so complex and uncertain that the costs of forcing consolidation on unwilling parties, even where conceptually justifiable, appear to outweigh any benefits secured. When parties consent to consolidation, on the other hand, it can provide important benefits for resolving related cases and identical claims. Even limited forms of consolidation can offer concrete and meaningful benefits.

Forced Consolidation

The Federal Arbitration Act aims to advance the administration of justice, but adopts no particular view of the purposes of arbitration other than to enforce written agreements to arbitrate “in the manner provided for in such agreement.” This governing principle implies that parties may be required to use consolidation in arbitrations only when they have agreed to do so, and may expressly preclude its use. The U.S. Supreme Court has confirmed these conclusions in decisions concerning class actions that require proof that the parties to an arbitration agreed to the use of class actions, and that uphold their agreements to preclude the use of class actions.

While these principles are clear in theory, they are difficult to apply in practice. With regard to the standard for proving consent, for example, Stolt-Nielsen v. Animal Feeds International, 559 U.S. 662 (2010), held (5-4) that an agreement to arbitrate could not alone be the basis for finding intent to permit class arbitration. In that case, however, the parties stipulated that they had reached no agreement on the use of class actions, so it remains to be seen what proof the courts will require to establish consent where the parties dispute the issue. To what extent, for example, will language in an agreement requiring the application of a particular set of rules be sufficient to establish consent, or the lack of consent, to consolidation? Consent may reliably be inferred where an agreement explicitly adopts a set of rules that in turn explicitly allow consolidation. But uncertainties arise if arbitrators allow the use of consolidation in a manner that goes beyond the terms of a rule or on the basis of general objectives in a set of rules, such as efficiency or justice. Is it correct to infer no consent when the rules adopted by the parties have no consolidation provision?

Additional uncertainty results where agreements to arbitrate contain provisions that imply consent or lack of consent to consolidation. For example, discovery limitations (e.g., on the number of depositions), limits on the number or types of witnesses, a very limited number of hearing days, or the presence of confidentiality requirements may suggest that the parties contemplated individual hearings, especially when the consolidation sought is of scores or hundreds of claims.

Similarly, although the court has held in American Express v. Italian Colors Restaurant, 570 U.S. __, 133 S. Ct. 2304 (2013), and AT&T Mobility v. Concepcion, 563 U.S. 321 (2011), that provisions in arbitration agreements precluding the use of class actions must be enforced despite determinations that they are unconscionable under state law, or that individual claims would not be viable, such prohibitions may be unclear or less comprehensive than those present in the Amex and AT&T cases.

These uncertainties as to the standards for determining consent are greatly magnified by the fact that parties may choose to assign the decision on consolidation to arbitrators. Given the very limited judicial review of arbitrator decisions set out in Oxford Health Plans v. Sutter, 569 U.S. __, 133 S. Ct. 2064 (2013), arbitrators therefore can make decisions on consolidation that are inconsistent with the court's standards.

Furthermore, the court has yet to make clear whether the evidence needed to establish that the parties to an arbitration agreement assigned responsibility for deciding whether they consented to consolidation must meet the “clear and unmistakable” standard of First Options of Chicago v. Kaplan, 514 U.S. 938 (1995), rather than the generally applicable state-law principles that govern the formation of contracts. The court's distinction between substantive and procedural arbitrability in First Options, modified thereafter in Howsam v. Dean Witter Reynolds, 537 U.S. 79 (2002), led lower federal courts overwhelmingly to conclude that whether an arbitration agreement authorized class actions was a “substantive” issue rather than one that could be seen as “procedural.” But in Green Tree Financial v. Bazzle, 539 U.S. 444 (2003), four justices declined to apply the presumption in favor of judicial determination of arbitrability to the question whether an arbitration agreement prohibited the use of class actions. That issue, the justices explained, “concerns neither the validity of the arbitration clause nor its applicability to the underlying dispute between the parties. ' Rather the relevant question here is what kind of arbitration proceeding the parties agreed to.” Id. at 452.

The plurality opinion in Green Tree led arbitration services and arbitrators to conclude they could decide (without meaningful judicial review) whether parties had agreed to allow the use of class actions. Over 200 class actions were commenced, including some 100 under an AAA program. See, P. Christine Deruelle & Corey Berman, “The Future of Class Arbitration: Recent Supreme Court Jurisprudence on Arbitration Clause Interpretation and Enforceability,” Weil Class Action Monitor, p.9, n.9 (July 23, 2013).

Simultaneously, at least five federal circuit courts held that arbitrators could also decide whether the parties agreed to allow the use of consolidation (See, Robert W. DiUbaldo, “Evolving Issues in Reinsurance Disputes: the Power of Arbitrators,” 35 Fordham Urb. L.J. 83, 88 n.32 (2008)), leading to widespread efforts to consolidate cases or claims between 2003 and 2010. Most of the class action activity triggered by Bazzle ended after the Supreme Court decided in Stolt-Neilsen that class actions could not be imposed on parties with arbitration agreements silent on that issue, but efforts to impose (or find consent for) consolidation have continued, without authoritative resolution other than the very clear rule that a very clear delegation of the issue to arbitration is subject to very limited review even of very clearly wrong decisions.

Additional risks and uncertainty that arise when arbitrators insist on imposing consolidation include the possible effects of state laws based on the Revised Uniform Arbitration Act. The Act provides authority for courts to order consolidation on grounds similar to those that permit consolidation in litigation where the agreement to arbitrate does not prohibit consolidation. See, Section 10 (2000), reprinted in Stephen K. Huber & Maureen A. Weston, Arbitration: Cases and Materials, p. 491 (3d Ed. 2011) (LexisNexis). Do these statutes preclude arbitrators from consolidating claims, and if so for all purposes or only for the purpose of joint decision?

Another problem is reflected in the Sixth Circuit Court of Appeals' recent decision allowing a subcontractor to sue design professionals despite having participated in a consolidated arbitration that included the issues on which the subcontractor based its subsequent claim. See, W.J. O'Neil Co. v. Shepley, Bullfinch, Richardson & Abbott, No. 12-2320, 2014 U.S. App. LEXIS 16607 (6th Cir. Aug. 28, 2014). Will res judicata (i.e., claim preclusion) lose some of its effectiveness as a method for securing finality with regard to parties who are compelled to participate in consolidated arbitration? Several Justices have expressed doubt as to the propriety of requiring parties forced into class action arbitrations to opt out rather than requiring that they be included only if they opt in. So, even where consolidation may be properly mandated for a set of related claims, it may nonetheless be inapplicable to parties that have not consented to arbitrate those claims.

However well intended, therefore, efforts to mandate consolidation in arbitration without adequate proof of consent have been, and are likely to continue to be, little if any more productive than efforts to mandate nonconsensual class actions. Multi-party procedures are likely to prove useful in arbitration only through consent.

Advantages of Consent

What seems most important about all this, however, is not that consolidation has no future in arbitration. Rather, the principal lesson is that arbitrators and parties alike need to stop attempting to mandate full-fledged consolidation without proof of consent, and should instead exploit more systematically the potential benefits of using limited, pre-hearing consolidation based on actual consent. Consolidating arbitrations for all purposes, including hearing and decision, has a greater impact than even class actions in that they allow no right for individual parties to opt out. But unlike in class actions, consolidation can be used in arbitration for limited purposes. Arbitrations can be consolidated for general administration (thereby reducing the fees charged by arbitration services and enabling a single panel to administer the prehearing process); for document discovery (e.g., to compile a set of documents applicable to all the consolidated cases); for interrogatories (by making the evidence obtained usable in all proceedings in which it is admissible); for deposition discovery (e.g., to avoid multiple depositions of the same witness on the same issues); for some or all pre-hearing motions; and to permit the use by all parties in multiple proceedings of the same briefs and written witness statements. A similar, though mandatory, process is available in the federal courts through multi-district litigation designations. Finally, arbitrators can, in lieu of consolidating hearings, designate, hear and decide one or more typical claims rather than allowing the entire set of related matters to remain in limbo while the parties litigate whether all the relevant claims should be consolidated for all purposes.

Some of these processes may be undertaken without risk of serious challenge. A party opposed to consolidation for some purposes, moreover, may accept joint processing for more limited purposes to avoid unnecessary costs and duplication of effort. Deciding individual, representative claims could also significantly expedite the resolution of other, related claims. Parties in litigation faced with numerous, related claims in particular areas of commercial activity (such as construction, reinsurance, and franchising) sometimes designate specific cases for trial (or mock trial) in order to provide concrete indications of likely results, which they then use as guidance in settling or streamlining for trial the remaining claims. The same, rational process is likely to play out in arbitration; it is less dramatic than forcing joint hearings, but more effective at delivering concrete and useful results. This process is most likely to be effective in sets of claims that are each substantial enough to justify separate hearings. But it is also likely to be helpful in sets of claims that are each too insubstantial to warrant separate hearings, since joint processing and the decision of even one or a few such claims can lead to practical resolutions aimed at avoiding costs (including the possibility of attorneys' fees), the possible application of collateral estoppel, and the loss of commercial goodwill.

Finally, a danger of inconsistent rulings exists where different arbitrators are making decisions on the same issues. But few situations have been identified in discussions about consolidation in which parties have had to live, against their will, with inconsistent outcomes caused by inconsistent decisions of courts or arbitrators due to the unavailability of consolidation. The danger of inconsistent results is largely theoretical. Where inconsistent results actually occur, they may well reflect the fact that the correct outcome in those cases is itself uncertain, rather than any lack of competence or legal discipline on the part of the arbitrators or judges involved. Inconsistent results may well be no more than a signal to the parties of their need for realistic compromise.

Adopting a practical, consensus-based approach to consolidation in arbitration will not relieve arbitrators of their duty to decide legal issues. In some situations, parties may insist on arbitrator rulings on consolidation-related issues, and some may refuse to agree on any practical steps. In the absence of consent, however, arbitrators will be on firmer ground in requiring practical steps that do not prejudice the right to separate hearings and decisions than in issuing rulings that purport to impose the most consequential aspects of consolidation. The law is simply too uncertain to justify confidence in almost any one-sided determination. Even where the law is clear, such decisions are likely to lead to drawn-out, costly litigation during which all progress on the processing and resolution of the claims at issue is suspended. A flexible, consensual use of collective processing, rather than formal consolidation, is something arbitrators should seriously consider, regardless of their views on the legal merits of consolidation.

Finally, it is appropriate to note that the interest among parties and arbitrators in utilizing class actions and consolidation in arbitration reflects the continuing, potential utility of procedures adopted in the Federal Rules of Civil Procedure on the basis of some 150 years of litigation experience. The Federal Rules have real advantages in the sorts of litigation for which they are designed, whether conducted in court or in arbitration. They include provisions dealing with class actions and consolidation (though not with the issue of effects on parties that have not agreed to arbitrate). They are inapposite in contexts dissimilar to conventional federal litigation. But they are useful in the arbitration of disputes similar, if not identical, to those that arise in the federal courts. It should not be surprising that parties to such disputes that want to use arbitration for its many unique advantages (such as the choice of judges, applicable law, venue, confidentiality, etc.), may also want to utilize the Federal Rules where they appear advantageous compared to conventional arbitration procedures. (Federal Arbitration, of which the author is a founder and chairman, uses the Federal Rules of Civil Procedure as its default rules of arbitration, subject to appropriate modifications and to changes made by the parties.)

Abraham D. Sofaer is the George P. Shultz Senior Fellow at Stanford University's Hoover Institution and a founder and chairman of Federal Arbitration, Inc. He previously served as professor of law at Columbia University and as U.S. District Judge for the Southern District of New York. This article also appeared in the New York Law Journal, a sister publication to this newsletter.