I'm sure that what I am about to describe has happened to many advocates in international arbitration proceedings. You have spent hours preparing for the cross-examination of a witness. You have planned your questions with surgical precision to extract only the damaging admissions the witness must make. You have taken pains to ensure that you don't ask the infamous one-question-too-many. You have deliberately avoided the question "why." You start the examination, and just as you are hitting your stride, an arbitrator jumps in and asks questions you deliberately chose not to ask, undermining all your careful planning. You sit there with your best poker face and remember the words of that great philosopher Mike Tyson: "Everybody has a plan until they get punched in the mouth."

Here's an example from a real case in which I was one of the arbitrators. Mr. X was an employee of a party to the arbitration (let's call it "Smith"). He was intimately involved in the events that were the subject of the arbitration, and had written, received, or been copied on a majority of the emails relating to those events. Despite this, he never submitted a witness statement, and so never appeared at the hearings. His absence was conspicuous. Smith's lawyer never explained to the Tribunal the reason for his absence. The lawyer for the opposing party (let's call it "Jones"), decided to make an issue of the missing Mr. X for the first time at the hearings, with the apparent aim of asking the arbitrators to draw an adverse inference from his absence.

During his cross-examination of Ms. Y, who, like Mr. X, was also employed by Smith, Jones's lawyer established that Mr. X was involved in the events that gave rise to the dispute, that he was still employed by Smith, and that he was currently at work. The lawyer then cut off Ms. Y's attempts to offer any explanation of why Mr. X had not submitted a witness statement. One of my fellow arbitrators then stepped in and asked Ms. Y the obvious question: Why had Mr. X not submitted a witness statement? Her answer: At the time the witness statements were being prepared, Mr. X was recovering from serious injuries sustained in a car accident, and had only recently recovered sufficiently to return to work.

This fact might have come out on redirect examination, but not necessarily. The lawyer conducting the redirect might not have known the answer, and so might have feared that a question would do more harm than good. In any event, the central point is that there are many cases in which facts come out or legal arguments are advanced as a result of arbitrators' questions that otherwise might not see the light of day.

All this is a long way of introducing the topic I want to write about in this column: arbitrators' questions. I think most readers would agree that in the circumstances described above the arbitrator's question was perfectly appropriate. If the arbitrators were going to be asked to draw an adverse inference from a witness's absence, it was essential for them to know whether there was a good reason for it. And if a lawyer questioning a witness dances around the issue, but fails to ask the only question that matters, in my view, it is not simply permitted, but perfectly proper for an arbitrators ask the question.

Significant Assistance

Most readers, I think, would agree that an arbitrator can ask the questions that, in her discretion, she deems necessary in order to understand and decide the case. Having said this, however, arbitrators' questions can be problematic. This is because they can sometimes end up helping one side to the dispute at the expense of another. This is clear in a way that was unlikely to be outcome-determinative in the case involving Mr. X. But sometimes an arbitrator's question may bring to light facts that may be decisive.

Take the following example from a real case involving a contract governed by New York law. In an arbitration proceeding between Smith and Jones, Smith took the position that the contractual language on which the dispute turned should be construed strictly according to its plain terms and that, so construed, it should prevail. Jones did not seriously dispute that the contract, read strictly, favored the interpretation advanced by Smith. Rather it objected that this interpretation had commercially unreasonable results and it argued, based on New York law, that the Tribunal should reject Smith's interpretation on that ground. Neither party had submitted any evidence regarding which party had drafted the pertinent language. Then, toward the end of the hearings, one arbitrator asked a witness from Jones: Which party drafted the disputed language? Jones's witness answered: We did. While the arbitrator had no idea that would be the answer, that testimony proved to be of significant assistance to Smith. Jones's argument that the tribunal should disregard the strict language of a contract—that it did not seriously dispute favored Smith—was less persuasive once Jones had copped to the fact that it had drafted the language in question.

Limitations?

This raises the issue of whether there is any limit on arbitrators' questions. While most would agree that it is perfectly appropriate, as a general matter, for arbitrators to ask questions, can arbitrators' questions ever cross a line? And, if so, how do we identify that line?

I want to distinguish (i) the biased arbitrator who asks questions in order to assist her appointing party; and (ii) the diligent arbitrator who respects her obligation of neutrality, but to whom questions occur during the course of the case.

This article is not about the first situation. Arbitrator bias is a topic in its own right, and an arbitrator asking questions to assist her appointing party is simply one way she might display bias. The simple answer to the biased arbitrator is this: Stop it.

I want to address the second situation—the good faith arbitrator to whom questions occur that may bear on the outcome of the case. Are there situations in which that arbitrator should bite her tongue and not ask those questions? This issue requires special consideration where the contemplated question is binary: it admits only of two contradictory answers, one of which helps one party and hurts the other.

Consider an example. Imagine an arbitration clause where the parties have agreed that any dispute should be resolved in accordance with the Rules of the International Centre for Dispute Resolution (ICDR), the international arm of the American Arbitration Association. The ICDR Rules provide in Article 31(5) that, unless the parties agree otherwise, "the parties expressly waive and forego" any right to punitive damages. Smith nonetheless requests punitive damages. Jones raises no objection to that request based on the ICDR Rule. The arbitrator is aware of the ICDR Rule prohibiting the award of punitive damages and wonders why it should not apply in this case.

Since the parties agreed to conduct their arbitration under the ICDR Rules, an arbitrator has a duty to apply those rules. While I can see an argument the other way, in my view, Article 31(5) is best read as a prohibition on an arbitrator's authority to award punitive damages and applies whether or not a party raises that rule. Even though Jones did not raise that rule, in my view there would be nothing inappropriate in the arbitrator asking Smith why that rule does not prohibit the award of punitive damages.

Moreover, when arbitrators believe that an issue in a case (e.g., should they award punitive damages?) is best resolved by reference to a ground raised for the first time by the arbitrators (the prohibition in Article 31(5) on the award of punitive damages), they must give the parties an opportunity to comment. Failure to do would give rise to a legitimate basis to challenge the award on the ground that the losing party, having not been permitted to comment on the basis of the arbitrators' decision, did not have a fair opportunity to present its case. See, e.g., Carribean Niquel v. Sté Overseas Mining Investments, Arrêt No. 785, 29 juin 2011 (1-23321) première chambre civile (vacatur of award by Cour de Cassation (France's highest civil court) because the arbitrators relied on a damages theory not advanced by either party).

I think the same conclusion would be warranted where Smith sought consequential damages in a case where they were contractually prohibited, but where Jones had failed to raise that contractual bar. An arbitrator has a duty to resolve the case by reference to any applicable contract, and, thus, there would be nothing improper in her, sua sponte, raising the contractual bar on consequential damages. See, e.g. ICC Rule 21(2) ("The arbitral tribunal shall take into account the provisions of the contract… " (emphasis added).)

But consider a different example. Suppose Smith brings a claim against Jones for breach of contract in circumstances where that claim is time-barred under the applicable statute of limitations, which is that of New York. Jones, however, fails to raise a limitations defense. The diligent arbitrator notices what Jones does not: the claim appears, on the face of the pleadings, to be time-barred under New York law. Should the arbitrator raise this issue? This question can be broken down into two separate ones. Is an arbitrator required to raise it? If not, is she permitted to raise it?

Under New York law, the statute of limitations is an affirmative defense and, as such, can be waived if not raised. Mendez v. Steen Trucking, Inc., 254 A.D.2d 715, (App. Div. 1998). In fact, in New York, a court is not permitted to take judicial notice sua sponte that an action is barred by the statute of limitations. Id. As a result, it is reasonable to conclude that an arbitrator is not required to raise it. The issue does not implicate arbitral authority; if an arbitrator issues an award finding a breach of contract in circumstances where the claim for breach was time-barred, but no limitations defense had been raised, she has not exceeded her authority or created any other basis to challenge an award.

But if she is not required to raise it, is she permitted to? As an initial matter, it is worth approaching this question by noting that it is doubtful, absent evidence of bias, that an arbitrator raising such question would have created grounds for a successful challenge to an award. For example, in Cellu-Beep Inc. v. Telecorp Communications, Inc., 2014 WL 358551 (SDNY), the respondent had made a motion to dismiss in an arbitration proceeding, but had not invoked the statute of limitations. In a subsequent call with the parties, however, the arbitrator, "unprompted by anything the parties said," asked the respondent whether it would also be moving on the basis of statute of limitations. The respondent then did so in its reply brief, and the arbitrator later issued an award dismissing the case on grounds of statute of limitations. Noting that the standards governing the impartiality of arbitrators are less stringent than those for judges, the court rejected a challenge to the award based on evident partiality under section 10(a)(2) of the Federal Arbitration Act.

However, even though an arbtirator's sua sponte question about the statute of limitations is unlikely to create a basis to challenge the award, there is nonetheless something problematic about that question. By her question, the arbitrator assisted one side to the case by bringing to its attention an outcome-determinative argument that it otherwise would have waived but for the arbitrator raising it.

The issue raised here is a fundamental one: is an arbitrator's obligation of independence and impartiality coterminous with the narrow circumstances in which an arbitral award can be vacated on grounds of bias? Or to ask it another way, is it the case that, as long as an arbitrator does nothing that creates a successful basis to challenge an award, she is permitted to ask any question, even one she knows in advance is likely to assist one party and hurt the other?

Reasonable arbitrators will likely disagree about the right approach here. Some, no doubt, will take the view that as long as an arbitrator does nothing that undermines the enforceability of any award, anything goes. I would argue against that categorical position. It is important not only that an award is enforceable, but also that the parties have confidence in the process, feel that they were treated fairly even when the perceived unfair treatment does not rise to the level of a ground to challenge an award. In a case where both sides are represented by able counsel, an arbitrator should tread cautiously when she, through a question, might suggest an outcome-determinative argument to a party that it would otherwise have waived.

Conclusion

I want to end by discussing the situation which prompted me to think about the issue addressed by this article. About a year ago, I was a party-appointed arbitrator in a case where my counterpart came up with a creative argument that favored his appointing-party which that party had not raised. While space limitations (and confidentiality obligations) make it impossible for me to go into detail, the key point to stress was that the argument advanced by the arbitrator was not straightforward, such as his noticing from the face of a pleading that a claim was time-barred or from the face of the contract that it prohibited certain relief. Rather, it was a sophisticated argument that rested on reading several seemingly unrelated provisions of the underlying contract together in an inventive way. The arbitrator advanced this argument through a series of questions to counsel for his appointing-party. When it became apparent what was going on, the opposing lawyer became very agitated and vociferously expressed his concern that the arbitrator was unfairly assisting his appointing party. I found myself quite perturbed, as did the presiding arbitrator, who cut off the questioning and took a break, during which he had some quite strong words for the arbitrator. Even assuming the arbitrator had not been animated by bias, in my view, he should nonetheless have held his tongue. His questioning had the effect of assisting one party to the case at the expense of another by suggesting a non-obvious argument that would never have been advanced had he not raised it, and that undermined the opposing party's confidence in the process.

John Fellas is a partner at Hughes Hubbard & Reed.