2002 Freshfields Lecture.
19 Arb. Int’l 279 (2003).
Arbitration’s Protean Nature:
The Value of Rules and the Risks of Discretion
by WILLIAM W. PARK*
“In most matters it is more important that the applicable rule of law be settled than that it be settled right.” Louis Brandeis1
The Freshfields Lecture for 2002 questions the wisdom of unfettered arbitrator discretion. The author suggests that the absence of procedural constraint on the arbitral tribunal can create more problems than it solves, often giving the impression of an “ad hoc justice” that damages the perceived legitimacy of the dispute resolution process. Challenging the prevailing orthodoxy about the costs and benefits of discretion, the Lecture explores the feasibility of including, in institutional arbitration provisions, a set of more precise procedural protocols in institutional provisions, to apply unless the litigants explicitly opt out of the default norms.
INTRODUCTION: THE WHY AND HOW OF ARBITRATION
A. Diversity of Motive and Method
LET US go back three-quarters of a century. In June 1927, the National Geographic Magazine published an article describing law reform under a Manchu emperor who reigned in the early eighteenth century. Emperor Kang-hsi decided that courts should be as bad as possible so his subjects would settle disputes by arbitration. Responding to a petition about judicial corruption, he decreed as follows:
Lawsuits would tend to increase to a frightful extent if people were not afraid of the tribunals. ... I desire therefore that those who have recourse to the courts should be treated without any pity and in such a manner that they shall be disgusted with the law and tremble to appear before a magistrate. In this manner ... good citizens who may have difficulties among themselves will settle them like brothers by referring to the arbitration of some old man or the mayor of the commune. As for those who are troublesome, obstinate and quarrelsome, let them be ruined in the law courts.2
So here is one reason to arbitrate: the hope of avoiding a grossly mismanaged judicial system. But there are other motives. For international transactions, arbitration offers the hope of reducing bias and the prospect of parallel lawsuits in different countries. There may also be the expectation (whether warranted or not) of confidentiality and expertise. In some countries, such as the United States, arbitration has been fuelled by a hope of keeping consumer and employment cases away from sympathetic civil juries inclined to award high punitive damages.
Arbitrations also show enormous variation in the mechanisms used to establish the facts and the law. A letter of credit dispute might be arbitrated in a few hours on the basis of documents only. At the other extreme, a large construction case could involve years of proceedings, with pre-trial discovery, depositions, motions on applicable law and jurisdiction, as well as witness statements and extensive cross-examination.
This variety should not be surprising, since arbitration (like dispute resolution in general) runs the gamut from large investment controversies to small credit card debt collection. The spectrum of subject matter includes construction, baseball salaries, biotech licences, expropriation, joint ventures, auto franchises, distribution and agency contracts, employment discrimination, insurance, collective bargaining agreements and Internet domain name disputes.
The moral flavour of arbitration differs dramatically from context to context. The values of fairness and efficiency that commend arbitration to sophisticated business managers often serve to condemn the process in consumer cases, where an arbitration clause might require an ill-informed individual to seek uncertain remedies at an inaccessible venue.
Finally, the public image and aura of arbitration will vary depending on perspective. In Western Europe, arbitration traditionally took the moral high ground, portrayed as an exercise in self-governance by the commercial community involving co-operation between the private sector (which conducted the arbitration) and the state (which enforced the award). In cross-border commerce, arbitration also was seen as providing a way for companies from different parts of the world to level the procedural playing field.
In developing nations, however, arbitration has often been perceived in a much less glorious light, as a process whereby secret tribunals undermine national sovereignty and legitimate governmental regulations. Ironically, the latter view has recently gained currency among certain segments of the US population disturbed about NAFTA investment claims brought by Canadians.3
B. Common Themes
Yet, notwithstanding its chameleon-like character, arbitration maintains a core essence. Litigants renounce the jurisdiction of otherwise competent courts in favour of a private and binding dispute resolution mechanism. Arbitration institutions usually purport to promote equal treatment and basic notions of fairness. Arbitrators are expected to possess integrity, experience, and the ability to be both a good listener and a careful reader. In most cases, litigants also want their arbitrator to be intelligent, although at least one case comes to mind in which a lawyer sought to disqualify an arbitrator whose strong intellect made it unlikely that the lawyer’s client would succeed with its clever but spurious arguments.
The interaction of arbitration’s diversity of form and unity of essence brings to mind the elusive Greek sea-god Proteus, who had the gift of altering shape while his substance remained the same. Similarly, arbitration is constantly reinventing itself to adapt to each particular case and legal culture, while retaining a vital core which aims at final and impartial resolution of controversies outside national judicial systems.
I. ARBITRAL DISCRETION
A. The Benefits of Procedural Autonomy
One reaction to arbitration’s protean nature has been an emphasis on broad grants of procedural discretion to the arbitrators. Arbitrators can conduct proceedings in almost any manner they deem best, as long as they respect the arbitral mission and accord the type of fundamental fairness usually called ‘due process’ in the United States and ‘natural justice’ in Britain, which includes both freedom from bias and allowing each side an equal right to be heard. Consulting the entrails of a disembowelled chicken might perhaps be off limits. Negative attitudes about augury aside, however, very few constraints limit the manner in which arbitrators go about their jobs.4
The absence of precise procedural rules is said to constitute arbitration’s strength, by allowing creation of norms appropriate to the contours of each dispute. Established dogma teaches that much of arbitration’s genius lies in giving carefully chosen individuals the freedom to apply just the right touch of this or that procedural principle – the je ne sais quoi of justice that leads to innovative and clever compromises.5 Like a bespoke tailor, the creative arbitrator cuts the procedural cloth to fit the particularities of each contest, rather than forcing all cases into the type of ill-fitting off-the-rack litigation garment found in national courts. While not totally false, this view is incomplete, as we shall see in a moment.
B. Two Meanings of “Rules”
This discretionary power exists not only in ad hoc proceedings, but also when the parties agree to a set of prefabricated institutional provisions, such as those of the International Chamber of Commerce or the American Arbitration Association. Here we encounter a slight linguistic challenge. In arbitration, the term ‘rules’ can bear at least two different meanings. First, there are stipulated frameworks of pre-set provisions (like the ICC or the AAA arbitration rules) that address matters related to the appointment of arbitrators and basic requirements of initial filings.6 Secondly, specific directives for conduct of the proceedings (in both fact-finding and legal argument) govern matters such as privilege and document production.
For better or for worse, rules in the first sense (prefabricated provisions) contain few rules in the second sense (canons for conduct of the proceedings), but leave the latter questions to the arbitrators. For example, the ICC Rules provide simply that the arbitrator may establish the facts by ‘all appropriate means’.7 Both the UNCITRAL and the AAA International Rules say that the tribunal may conduct the arbitration in ‘whatever manner it considers appropriate’.8 Even the LCIA Rules (which do a better job than most in transforming litigation practice into precise directives)9 are explicit in giving the arbitral tribunal the ‘widest discretion to discharge its duties’.10
The same grant of discretion is found in modern arbitration statutes.11 Enlightened arbitration statutes today usually limit mandatory judicial review to matters such as bias, excess of authority and gross procedural irregularity.12 In some cases they may also admonish the arbitrator to act with fairness and to adopt procedures suitable to the circumstances of the particular case.13
II. THE DOWN SIDE OF DISCRETION
A. The Need for Default Procedural Protocols
The time has come to present this article’s tentative thesis, which with some simplification might be presented as follows: the benefits of arbitrator discretion are overrated; flexibility is not an unalloyed good; and arbitration’s malleability often comes at an unjustifiable cost.14 Therefore arbitral institutions should give serious consideration to adopting provisions with more precise procedural protocols to serve as default settings for the way arbitrations should actually be conducted. These directives would explicitly address questions such as documentary discovery, privilege, witness statements, order of memorials, allocation of hearing time, burden of proof and the extent of oral testimony.
Unrealistic expectations are resentments waiting to happen.15 Arbitration is no exception. When an arbitrator adopts a model of procedural fairness different from what was anticipated by one of the parties, the arbitrator may well believe that his or her approach is ‘the usual way things are done’. In an international context, however, competing experiences will almost always be available to indicate that other approaches are not uncommon. While a page of history is certainly worth a volume of logic,16 arbitration’s enormous variety means that even the best and the brightest may be reading from quite different pages — or may read from one page one day and another the next. An arbitrator might say, ‘In my experience it is common to allow two rounds of briefs, with the respondent having the last word’. Yet on another day, with a slight adjustment in phraseology, the same arbitrator could assert the contrary, that it is ‘not uncommon’ to give the claimant the last word — and both statements would be correct.
The issue is not whether a model exists for a particular decision, but why one paradigm rather than another should prevail. While it seems almost axiomatic that parties themselves should be free to fashion their arbitration as they see fit, it is less evident that arbitrators ought to be in the business of setting norms for specific procedural questions on an ad hoc basis. For they may already have seen which side will be advantaged by one rule or the other.
This may not matter when all parties share or have adopted a common legal culture,17 or belong to a relatively homogeneous community that shares confidence in the individuals chosen to decide the case. Two Boston law firms arbitrating before a well-known Boston arbitrator would normally be expected to behave professionally and accept rulings that comport with their common range of expectations on matters such as witness sequestration and document production.
However, if backgrounds and experiences differ materially, the ad hoc imposition of procedures uncustomary to one side and not announced in advance, risks reducing the perception of arbitration’s legitimacy.18 The aggrieved party may then feel justified attempting to disrupt and derail the proceedings with charges of procedural unfairness.
B. Arbitral Orthodoxy
To many in the arbitration community, any suggestion that arbitral discretion should be curtailed may be as welcome as ants at a picnic.19 The flexibility inherent in arbitrator discretion not only constitutes a pillar of orthodoxy,20 but rests on deeply entrenched practical considerations. Arbitral institutions that aspire to market their services globally are understandably shy about taking sides in long-standing debates between different national legal systems, particularly on those controversies that divide continental and Anglo-American civil litigation.21 By leaving procedural matters to the arbitrators’ discretion, institutions side-step the hard choices about what exactly it means to conduct a fair and efficient proceeding.
Consequently, one must recognise the enormous conceptual and practical problems attached to any suggestion of reduced arbitral discretion. More than once, the dilemmas of this topic have brought to mind the notice hanging at the entrance to Dante’s Inferno: Lasciate ogni speranza, voi che entrate — ‘Abandon all hope, you who enter here’. 22 Better minds than mine have sunk beneath the waves trying to resolve the tension between rules and discretion. As our French colleagues say, Je m’interroge. Thus, this article remains very much a work in progress, and invites the audience’s merciless critique.
III. ARBITRATION’S ARCHITECTURE
A. Institutional Provisions
Let us return briefly to the architecture of arbitration. The American Arbitration Association lists over 30 different sets of arbitration provisions, with special procedures for securities transactions, finance, construction, collective bargaining, patents and employment disputes, as well as real estate transactions in Hawaii and Michigan and auto accidents in New York. On this side of the Atlantic, a visit to the Internet yields websites for at least a dozen London-based organisations devoted primarily to arbitration, and no less than 25 professional academies and trade associations that purport to sponsor arbitration as part of a more general mission. And there is no dearth of provisions developed by continental-based institutions and by Chambers of Commerce throughout the world.
In some cases these provisions do differ. For example, the AAA International Rules restrict punitive damages, while the domestic rules do not; and the LCIA Rules give the tribunal a clear right to join consenting third parties, which is not present in most other procedural frameworks.23
More often than not, however, these provisions are remarkably similar. Assuming that each side gets an opportunity to be heard, conduct of the proceedings is left to the arbitrators, permitting arbitral institutions to avoid the nitty-gritty procedural questions where the real demons lurk. Few pre-set arbitration provisions tell us whether pre-trial depositions are allowed, whether a party has a right to exclude one fact witness when another is testifying, or whether a log must be created to identify allegedly privileged documents withheld during discovery. Attempts at procedural precision usually involve not rules but ‘guidelines’ (such as the UNCITRAL Notes for Organizing Hearings)24 which the arbitrators are free to ignore — and often do.
B. Illustrative Questions: Privilege and Discovery
The dark side of all this discretion lies in the discomfort that a litigant may feel when arbitrators make up the rules as they go along, divorced from any precise procedural canons set in advance. Discretion may not be objectionable within a close-knit community (for example, among the diamond dealers in Amsterdam) or when everyone shares or accepts a common legal culture (as might happen in a construction arbitration where the lawyers and arbitrators share long-standing professional relationships).25 However, in a heterogeneous transaction with parties, lawyers and arbitrators from disparate places, anxiety rather than comfort may result from a level of arbitrator discretion that permits an arbitrator to make critical procedural decisions after he or she has sized up the parties and the controversy.26
To illustrate, imagine an arbitration between a Swiss company and a US corporation. One side requests pre-trial production of a memo created by in-house counsel. The other objects on the basis that the document is protected by privilege. What is the arbitrator to do?
In the United States, the attorney-client privilege generally applies equally to all lawyers, whether independent or employed by the client.27 By contrast, in Switzerland and many other countries, only communications to outside counsel are protected by professional secrecy.28
Equally troublesome is the very notion of pre-trial document production. In many parts of the world, including most of continental Europe, litigants do no more than give each other advance copies of the documents to be relied upon during the hearings, so as to avoid undue surprise. By contrast, in the United States a practice has developed by which the parties must produce to each other broad categories of dispute-related material, including documents that might help prove the adversary's case. Discovery serves as a vacuum cleaner to hoover up even marginally relevant pieces of paper that might lead to admissible evidence.29
How should the arbitrator choose between these divergent models of privilege and discovery? One approach would be to apply rules in a discriminatory manner, looking to the parties’ national practice and expectations. Accordingly, only communications to the US in-house lawyer would be privileged, and only the US side would be required to produce broad categories of documents.30
Instinctively, good arbitrators shrink from assigning procedural benefits and burdens according to the parties’ national practices. Giving one side a stark procedural handicap is an excellent way to invite challenge to an award.31 However, it can be perilous to decide which principles to follow after the proceedings have revealed the parties’ positions, indicating who will get the awkward side of a rule.
C. Consensus and Legal Culture
Some questions, of course, will in practice be resolved by a race to the lowest common denominator. For example, in international arbitration it is now generally expected that lawyers will prepare witnesses by discussing the case in pre-hearing interviews,32 even though many countries forbid this practice.33
On other questions, however, no consensus exists. For documentary discovery, continental lawyers generally feel deeply about what they perceive as the abusive US ‘fishing expeditions’ and scatter-gun tactics. Conversely, many lawyers in the United States believe that the bargain for arbitration never included renunciation of what for them is a basic right to shoot first and aim later.
In addressing such questions, we are all laden with baseline baggage that pushes us to presume our own conclusion, according to culturally-influenced assumptions built into our backgrounds.34 For example, two arbitrators might agree that there must be a reasonable ‘proportionality’ between the burden of producing a document and the document’s potential for enlightening the tribunal. Yet, in applying such a proportionality principle, a Paris avocat would normally start with assumptions quite different from those of a New York litigator.
The matter of costs presents another illustration of cultural blinders. European arbitrators often assume that in international arbitration, automatically the loser will pay some portion of the winner’s legal costs. In fact, major institutional rules stipulate simply that arbitrators have discretion to allocate attorneys’ fees, but do not suggest how that power should be exercised.35 In at least one significant part of the world commercial community (the United States), costs do not normally follow the event.
If such procedural questions are subject to recognised norms, a litigant that did not like a particular ruling could still recognise that either side might have got the short end of the stick. But in a mixed legal culture lacking a common procedural roadmap, the loser may feel not only the sting of defeat, but also a sense of injustice.36
IV. ALTERNATIVE PROCEDURAL MENUS
A. “Rules Light” and “Rules Rich”
Of course, the arbitral tribunal can always set forth a systematic set of rules in an initial procedural order, issued at the outset of the arbitration. Frequently, however, arbitrators may be tempted to keep the first order relatively simple, precisely to reduce the prospect of unnecessary wrangling among tribunal members, as well as litigants.
The more radical approach suggested today reverses the situation. Rather than a blank page to be completed by arbitrators, institutional provisions could contain specific protocols that the arbitrator would be required to apply unless modified by agreement of all parties.37 Under the current regime of arbitrator discretion, the litigants are like diners in a fancy restaurant with a menu that we might call ‘procedure light’, which allows the chef to feed them whatever he wants, as long as each gets the same meal. To force the chef to add or subtract a dish, the diners must do so by a common accord.
By contrast, a better approach would make the litigants like a couple who are served a fixed meal that includes soup, fish, meat, vegetables, potatoes, salad, pudding and savories — which is to say, rules on privilege, discovery, time allocation and the like. One might call this a ‘procedure heavy’ or ‘rules rich’ menu. To change the menu the diners would have to indicate jointly what they did not wish to order, or what they wanted to add.
The reason for reversing the way dinner is served derives from the fact that once the arbitration begins, litigants almost by definition are more like a bickering old couple than an amorous twosome, and thus may not agree on much. The arbitrator is left to make up rules as he or she goes along, with the potential consequence that one side may receive procedure never expected and never really bargained for.
The problematic aspects of drafting such ‘procedure heavy’ protocols are obvious. It would be hard work, representing a tricky compromise between different norms.
But is difficulty the same as impossibility? Perhaps the drafters of more specific rules will end up like Don Quixote, tilting at windmills. But then again, they might get lucky, and be able to emulate international co-operation in other areas of commercial law.
The International Bar Association Rules of Evidence provide a notable illustration of such successful cross-border compromise.38 In suggesting more precise norms than those contained in most institutional arbitration provisions, the IBA Evidence Rules enhance what might be called the ‘objectivisation’ of arbitration.39 For example, the Rules require that a request for document production must identify either particular documents or ‘a narrow and specific category of documents’ and describe how requested documents are ‘relevant and material’.40