On the occasion of the 2008 ICC Institute Annual Day devoted to documentary evidence, I provided an update on the work of the IBA Rules of Evidence Subcommittee, 1 which had been tasked with considering revisions to the 1999 version of the IBA Rules on the Taking of Evidence in International Commercial Arbitration (the "IBA Rules"). The question of evidentiary privilege is one of the most difficult and controversial issues presented in connection with documentary evidence and the IBA Rules.

This article posits that regardless of the approach ultimately taken by the IBA Rules Subcommittee, it will be fruitful to recall and reconsider the range of options presented with respect to evidentiary privilege in the context of future arbitrations. Ultimately, we stand at the crossroads in the context of each arbitration as we attempt to determine the most efficient, economical and just procedures for the taking of evidence in that situation. First, the IBA Rules may not be applicable to the case in question. Second, even if they are, Preamble 2 of the IBA Rules provides that they may be adopted in part or adapted to the particular circumstances of each arbitration. 2 Third, a one-size-fits-all solution may be less appropriate for evidentiary privilege than it is for other aspects of the taking of evidence given the complexity of the issue and the wide diversity in approaches.

First, the article will address the status quo, including the lack of best practices and the wide discretion granted to arbitrators under the IBA Rules with respect to evidentiary privilege. In particular, this section will address the framework for evidentiary privilege under the IBA Rules and the question of whether and to what extent rules of privilege can or must be applied under the IBA Rules as mandatory applicable law. [Page424:]

Second, the article will explore some of the reasons why evidentiary privilege has proven to be a particularly challenging issue, both to rule-makers and arbitrators.

Third, the article will describe certain options and choices relating to the formulation of default rules of evidentiary privilege.

1. The lack of clear best practices with respect to evidentiary privilege in international arbitration

Today there are no clear best practices regarding the recognition of evidentiary privileges in international arbitration. In 2006, the question was considered in the context of a special supplement on best practices with respect to document production in international arbitration, including issues of evidentiary privilege. The consensus that emerged was that generally Article 9 of the IBA Rules should be taken into account. 3 However, as will be seen, Article 9 of the 1999 version of the IBA Rules provides only scant guidance on (a) the substance and application of the rules to be applied and (b) how to determine them.

Article 3.5 of the IBA Rules provides that a party to whom a request to produce has been addressed may object in writing for any of the reasons set forth in Article 9.2 of the IBA Rules, which states that "the Arbitral Tribunal shall, at the request of a Party or on its own motion, exclude from evidence or production any document, statement, oral testimony or inspection for any of the following reasons". Article 9.2(b) of the IBA Rules provides for such exclusion by reason of "legal impediment or privilege under the legal or ethical rules determined by the arbitral tribunal to be applicable". Also relevant is Article 9.2(g) of the IBA Rules, which provides that evidence shall also be excluded for "considerations of fairness or equality of the Parties that the Arbitral Tribunal determines to be compelling".

a. Substance and application of evidentiary privilege in the 1999 IBA Rules

The only limitation on the substance of the rules of privilege appears to flow from Article 9.2(g) of the IBA Rules, which provides an additional ground for exclusion on the basis of equality and fairness. The 1999 Working Group stated in the Commentary to the 1999 IBA Rules that this provision was intended to cover the situation in which the national rules of privilege of one national legal system would not be considered privileged under the rules of the other. 4[Page425:]

Procedurally, however, it is not entirely clear how Article 9.2(g) of the IBA Rules comes into play. Logically, one can imagine the process working as follows: first, a party objects on grounds of privilege; second, the arbitral tribunal determines that the applicable rules of privilege are such that they would require production of the document; third, the arbitral tribunal then, sua sponte, raises its own objection to the production of that document based on the grounds that the application of the chosen rule of privilege would result in unfairness; and, fourth, the arbitral tribunal finds its own argument compelling and excludes the document from production, presumably based on Article 9.2(b) together with Article 9.2(g) grounds.

In practice, steps two and three are usually collapsed. Considering them separately, however, can also help one to differentiate analytically between the two most common approaches to determining the applicable privilege-the choice-of-law approach and the default or transnational rules approach.

As will be discussed in more detail below, under the choice-of-law approach, choice-of-law rules are applied to determine the applicable national law or laws of privilege applicable to particular documents or parties. A secondary, corrective ratcheting-up or ratcheting-down (MFN5 or LFN) is often necessary in the international arbitration context to prevent unfairness or inequality. In the transnational rules approach, a blended set of rules is applied in order to avoid unfairness in the first instance. In the end, the substance of the rules so identified or chosen may be quite similar. One of the central questions explored in this article is whether, for this or other reasons, it is legitimate and expedient to adopt the default or transnational rules approach.

i. Applicability of Article 9 of the IBA Rules to documents submitted

Another question regarding the application of the rules of evidentiary privilege in the 1999 IBA Rules relates to Article 3.1 of the IBA Rules, requiring parties to submit all documents upon which they rely. Article 3.5 of the IBA Rules, providing for objection on Article 9.2 grounds, appears to relate only to the "production" of documents in response to a request to produce, as opposed to the submission of "all documents available to it on which it relies" pursuant to Article 3.1. Is a party entitled to withhold documents on which it relies on the basis of legal impediment or privilege, or would be inherently unfair to allow a party to rely on a document it will [Page426:] not disclose? Should the decisionmaker order production, deeming the privilege to have been waived in such a case? In practice, this question is largely moot because parties are not usually held to the requirement in Article 3.1 that they submit all documents on which they rely. The concept of reliance is one that could also be clarified. Can the party avoid reliance, and thus production, by avoiding all reference to the document in its written submissions?

The better solution is probably to conclude that Article 9.2 of the IBA Rules does not require a specific triggering provision, i.e. that Article 3.5 need not extend to documents to be produced. This interpretation is supported by the fact that Article 9.2 also refers to the exclusion of statements, oral testimony or inspections. Neither Article 4 (Witnesses of Fact) nor Article 5 (Party-Appointed Experts) of the IBA Rules references Article 9 at all. Article 6 (Tribunal-Appointed Experts) references Article 9.2, but only in the context of the tribunal-appointed expert's power to request documents. Only Article 7 (On Site Inspection) provides a general conditioning of the admissibility of the site inspection on Article 9.2 grounds.

b. Choosing the applicable rules of privilege

The 1999 IBA Rules clearly state that the arbitral tribunal has discretion to determine the applicable rules of privilege. The only limitations on this discretion are contained in Article 2 of the IBA Rules. The following two subsections will address possible conflicts with mandatory applicable law and the "General Rules" that are contained in Articles 2.1 and 2.2 of the IBA Rules.

i. Article 2.1 of the IBA Rules: conflict with mandatory provisions of law

Article 2.1 of the IBA Rules provides that "the Rules shall govern the taking of evidence, except to the extent that any specific provision of them may be found to be in conflict with any mandatory provision of law determined to be applicable to the case by the Parties or by the Arbitral Tribunal." Having examined the content of the IBA Rules with respect to privilege, one can now ask (i) where conflicts with mandatory applicable law are likely to arise, (ii) whether a particular rule of privilege is in fact a "mandatory provision of law" and (iii) whether it should be deemed "applicable" in a particular case.

The laws and ethical rules respecting privilege, confidentiality and secrecy vary widely across jurisdictions. 6 The nature of these rules is important for determining the possible mandatory application in international arbitration, both in terms of the source of the rule and its characterization as substantive [Page427:] or procedural. Two general categories may be distinguished on the basis of the subject of the regulation: rules affecting the ability of a party to withhold evidence and rules affecting the ability of counsel to provide evidence in legal proceedings.

Rules affecting a party. Rules affecting the ability of a party to withhold evidence on grounds of privilege may be based in national rules of evidence, codes of civil or criminal procedure or the common law and are generally considered to be substantive in nature.

There can be no question that national rules of evidence and codes of civil and criminal procedure do not apply by their terms in international arbitration. However, the laws of many countries consider privilege to be substantive rather than procedural, even though they are often codified in rules of evidence. A US court will engage in a choice-of-law analysis to determine the applicable privilege between various US states or a US forum and a foreign country. But must an arbitral tribunal do the same? Hans Smit advises looking beyond this classification to the policy considerations behind the rule in taking decisions on the applicability of a particular national law in an international arbitration. 7

The principle that an arbitral tribunal in international arbitration proceedings has discretion to structure the proceedings as it sees fit is a common feature of the arbitration laws of most countries. Article 19 of the UNCITRAL Model Law on International Commercial Arbitration confers these powers in explicit terms:

"(1) Subject to the provisions of this Law, the parties are free to agree on the procedure to be followed by the arbitral tribunal in conducting the proceedings.

(2) Failing such agreement, the arbitral tribunal may, subject to the provisions of this Law, conduct the arbitration in such manner as it considers appropriate. The power conferred upon the arbitral tribunal includes the power to determine the admissibility, relevance, materiality and weight of any evidence."

Indeed, the determination of the "admissibility, relevance, materiality and weight" by its terms will include matters that are considered to be substantive in various jurisdictions. For instance, one could view the "right" to discovery or disclosure or the "right" to cross-examination as being equally substantive in nature and indeed mandatory under the rules of certain jurisdictions. However, international arbitrations without these features are common and are fully in accordance, for example, with the UNCITRAL Model Law. [Page428:]

When the IBA Rules refer to mandatory applicable law, this is best understood as a reference to the applicable lex arbitri. This interpretation is supported by the Commentary to the 1999 IBA Rules, which prefaced the discussion on Article 2 of the IBA Rules by stating that "[i]n addition, international arbitrations are subject to mandatory law relating to arbitration procedure at the seat of the arbitration". 8

In conclusion, Article 2.1 of the IBA Rules can be read to refer primarily to the applicable lex arbitri and not as a door through which national rules of privilege would become mandatory applicable law in international arbitration proceedings.

This is not to say that privilege should not be respected in some manner in international arbitration. The lex arbitri will require that the proceedings be fair and treat parties equally. This requirement is important with respect to the enforceability of an arbitral award and can also be derived from international instruments relating to the taking of evidence. 9 Thus, it may be necessary to apply rules of privilege in some form, but such application would not be in the direct manner anticipated by Article 2.1 of the IBA Rules in the case of "mandatory" applicable law, nor would it be inconsistent with the application of default or transnational rules of privilege, as will be discussed in more detail below.

Rules affecting counsel. Rules affecting the ability or duty of counsel to withhold evidence on the basis of privilege may be based in professional ethical rules, national rules of civil or criminal procedure or criminal laws. It is worth noting that a duty of confidentiality for attorneys appears to exist under the laws of most countries, including common-law countries. In many countries, however, including France and Spain, for example, disclosure of professional secrets may additionally subject counsel to criminal sanctions.

There are two major differences with respect to the mandatory nature of this category of rules affecting counsel as opposed to the rules discussed above affecting parties. Heitzmann points out that, in practice, the consequences of rules granting parties an evidentiary privilege and rules imposing a duty of secrecy on counsel are often treated similarly. 10 This may likely be so. However, the difference is important in considering the narrow question of whether such rules are "mandatory" within the meaning of Article 2.1 of the IBA Rules.

First, as a matter of logic, the question is less likely to arise in the context of rules affecting counsel than in the context of rules affecting parties for the simple reason that it is generally the parties who will submit or produce [Page429:] evidence in arbitral proceedings and not their counsel. One can imagine situations in which the testimony of a party's internal or external counsel would be relevant and material to a case, where it would be expedient to request the production of documents from such counsel, or cases in which the counsel itself is a party. However, in the concrete situation of the taking of evidence under the IBA Rules, objections on the basis of privilege will generally be made by a party on that party's behalf in relation to documents in its possession.

Second, if the duty of secrecy or confidentiality has been triggered, the arbitral tribunal cannot and should not ignore the practical consequences of counsel being asked to testify or provide evidence. The fact that attorney registration and regulation occurs at the national level can create complex situations in international arbitration, where counsels from various jurisdictions may participate in a hearing in yet a third country before an arbitral tribunal with no direct authority to enforce professional standards. 11 However, some jurisdictions make their ethical rules applicable extraterritorially, i.e. regardless of where the conduct of the lawyer occurs. 12 In such situations counsel may be subject to such rules in an international arbitration context. This concern is amplified with respect to attorneys admitted in or otherwise subject to the jurisdiction of national laws imposing criminal penalties for the disclosure of professional secrets.

Interestingly, both of the above cases appear to present a much closer call as to mandatory applicability in international arbitration than in the case of privilege rules affecting parties. It may be that, both in the case of ethical rules and criminal sanctions, the organizations or authorities promulgating those rules and laws intended to extend the reach of those rules extraterritorially. However, even this fact does not make their applicability mandatory in international arbitration. Criminal statutes will seldom come into play directly in international arbitration. Nonetheless, they may be relevant indirectly in making necessary determinations in an arbitration, for example with respect to the admissibility of evidence, the legality of an investment or the inability to testify or provide evidence.

Thus, as with rules relating to privilege held by parties, there is a compelling case for an arbitral tribunal to give effect to rules of confidentiality affecting counsel. The relevant laws or ethical rules may not be mandatory as to the arbitration, but they can be viewed as mandatory with respect to individual participants in that arbitration. [Page430:]

By contrast, it may well be that the duty of confidentiality could be considered a transnational rule. Most commentators draw the basic distinction between common-law jurisdictions with rules of evidentiary privilege and civil-law jurisdictions with a duty of secrecy. This overlooks the fact that common-law jurisdictions also impose a duty of confidentiality on practicing attorneys separate and apart from the privilege of clients to withhold documents associated with that advice. Although a general duty of confidentiality has not yet been codified in soft-law standards or international instruments, a survey such as the Lex Mundi survey documents the significant convergence of national legislation and professional rules on this point. 13

ii. Article 2.2 of the IBA Rules: conflict with "General Rules"

In considering the scope of the arbitrator's discretion in determining the applicable rules of privilege under Article 9.2(b) of the IBA Rules, one can now ask whether conflicts with the "General Rules" under Article 2.2 of the IBA Rules are to be expected. The IBA Rules define "General Rules" as including all other procedural rules according to which the parties are conducting their arbitration, including institutional and ad hoc rules.

In brief, to the extent that these rules do not address the question of privilege, there can be no conflict. As has been often noted in the literature, most institutional rules do not make specific reference to evidentiary privileges. There are, however, exceptions. Rule 32.2 of the ICSID Rules of Procedure for Arbitration Proceedings provides that "[t]he Tribunal shall for such cases establish procedures for the protection of proprietary or privileged information". The ICDR International Arbitration Rules provide in Article 20.6 that "[t]he tribunal shall take into account applicable principles of legal privilege, such as those involving the confidentiality of communications between a lawyer and client". However, because both of these provisions essentially mirror the requirements of Article 9 of the IBA Rules, no conflict arises.

c. Conclusion

There are currently no best practices with respect to evidentiary privilege in international arbitration, except a general consensus that privilege should be recognized in some manner. The 1999 IBA Rules provide a framework for parties to object to the taking of evidence on grounds of evidentiary privilege. However, they leave the substance of those rules largely to the discretion of the arbitrator. [Page431:]

The IBA Rules do not require or prescribe a conflict-of-laws approach as opposed to a default or transnational rules approach. The conclusion reached is that neither the IBA Rules nor mandatory applicable law nor the general rules require a particular method. Indeed, it has been shown that rules of privilege will generally not fall into the category of mandatory applicable law with respect to the arbitration. Exceptions hereto are ethical rules or criminal laws that may be applicable to counsel providing evidence in an arbitration.

Despite this lack of clarity regarding the substance of the rules of privilege applicable in international arbitration and the method of choosing them, there is considerable support for the proposition that evidentiary privilege should be preserved in some manner in international arbitration. This results from the obligation to preserve fairness and equality of arms, as it arises from various sources including, Article 9.2(g) of the IBA Rules, the applicable lex arbitri and international instruments relating to civil procedure and the taking of evidence. As will be seen below, arbitrators ought to give due deference to these principles in determining the nature and scope of the appropriate rules of privilege.

2. Challenges associated with rule-making in the area of evidentiary privilege

Privilege is probably the best example of a remaining area of evidentiary procedure subject to widely diverging practice in various jurisdictions and as to which no clear best practice or international consensus has yet emerged. Accordingly, it presents an important opportunity to expand the area of international convergence. 14 Some commentators have concluded that no convergence is possible due to the specific nature of the privilege issue and the extent to which it is tied to and regulated by national law. 15 Others have concluded that sufficient consensus has already emerged16 and/or that further convergence is a practical necessity. 17 This article takes the view that convergence is possible and that consensus will ultimately crystallize around default rules adopted either on an ad hoc basis in procedural orders or in the form of institutional rules, guidelines or protocols.

Evidentiary privilege is a particularly challenging issue due to the deeply-rooted controversies and uncertainties with respect to the relative merits of expanding the IBA Rules to cover privilege, the expected efficiency gains or losses as a result of expanded treatment of privilege in the IBA Rules and the appropriate level of procedural detail it is advisable to include with respect to privilege in the IBA Rules. Indeed, given the nature of these differences in [Page432:] opinion, evidentiary privilege is likely to remain a controversial issue, regardless of the approach ultimately adopted by the IBA Rules Subcommittee. In each case, it may be necessary to consider the substance, efficiency, fairness and scope of the applicable rules of evidentiary privilege chosen or identified by the arbitral tribunal or parties. The following section will attempt to describe the options available and the challenges associated with choosing among them.

a. Balancing economy and efficiency against fairness in the context of evidentiary privileges

In the case of legal impediment and privilege, the question arises as to whether creating new rules would draw additional attention to the issue and thereby increase the number of documents as to which privilege is asserted. This would in turn increase the amount of time the arbitral tribunal spends on deciding evidentiary challenges and decrease the efficiency of the proceedings.

As pragmatic and straightforward as economy and efficiency sound, it is difficult to prioritize these considerations in practice. First, economy and efficiency may not be ignored, but at the same time they may also not be prioritized to the extent that the fairness and justness of the arbitration is compromised. Second, although we can agree to general tendencies such as the likelihood that the availability of e-disclosure would tend to raise the costs of the taking of evidence, this will not be so in every case. For example, depending on the volume of documents, searching for the proverbial needle in the haystack can be quicker using electronic search tools than by physically reviewing the documents.

Additionally, the exclusion of evidence on the basis of legal impediment or privilege can result in a less efficient fact-finding or truth-finding procedure for the simple reason that it takes possibly relevant and material evidence outside of the consideration of the arbitral tribunal. However, the exclusion of evidence on the basis of privilege can also have the opposite effect of increasing fairness. If parties have relied on justifiable expectations of privilege, for example that their communications with counsel or settlement negotiations will remain confidential, excluding such evidence lends legitimacy and fairness to the proceedings. [Page433:]

A proper balance should be struck between (i) a party's right to assert privilege and (ii) the right of the other parties and the arbitral tribunal to have the benefit of that evidence. However, what constitutes the "proper" balance is susceptible to a number of views. Indeed, the views one holds regarding this balance appear to be closely related to one's basic beliefs about the nature of the truth-finding exercise in arbitration, legal training and individual experiences. 18

b. The value of default rules versus ad hoc determination on a case-by-case basis

On the one hand, more detail can simplify matters, especially for parties and arbitrators without extensive experience in a particular aspect of the taking of evidence in international arbitration.

As with efficiency and economy concerns, here too what may appear to be pragmatic questions actually relate to deeply held beliefs as to the ultimate value of flexibility and how to strike the proper balance. Indeed, there is profound disagreement as to the relative values of procedural certainty and arbitrator discretion, and the position one takes tends to be related to one's personal, legal and cultural background. William Park has argued memorably that "the benefits of arbitrator discretion are overrated; flexibility is not an unalloyed good; and arbitration's malleability often comes at an unjustifiable cost". 19 Experience shows that this may be a minority view, at least among European arbitrators.

With reference to evidentiary privileges, there are many factors that may be relevant to choosing the most appropriate rule and the necessary level of detail of that rule. As will be seen below, parties' expectations will vary according to their familiarity with and reliance upon evidentiary privileges in their home jurisdictions. Additionally, arbitrators may and do take different approaches to determining the applicable rules of privilege. Some may determine the applicable rules of privilege in a tailored manner through an exercise in choice-of-law and legal research. Others may be more comfortable with the freer exercise of discretion implicit in applying more generic or default rules of evidentiary privilege. As will be shown below, the substance of the rules determined as being applicable by these two very different approaches is not much different, although there are important ramifications to choosing the one or the other approach. [Page434:]

c. General tendency to avoid admissibility questions in international arbitration

A pragmatic approach widely practiced in international arbitration is to admit evidence freely and then take concerns that might have stood in the way of admissibility of that evidence in court proceedings into account in the assessment of the weight of that evidence. This approach may work well, for instance, with regard to hearsay evidence or authenticity questions. However, this approach is not especially helpful with respect to objections according to Article 9.2(b) of the IBA Rules that go to admissibility, including evidentiary privilege. Arbitrators should have a clear approach to evidentiary privilege from the start so that they are in a position to take decisions on evidentiary privilege before the evidence is admitted.

i. National admissibility rules do not apply

Absent party agreement, international arbitral tribunals are generally not subject to the rules of admissibility that constrain national courts. 20 Instead, arbitrators tend to err on the side of admitting potentially objectionable evidence, taking into account that the arbitral tribunal has discretion to weigh the evidence once it has been admitted. 21

This is partially attributable to the concern that parties be granted ample opportunity to be heard. Additionally, this broader concept of admissibility must be seen in relation to the civil-law notion of the responsibility of the arbitral tribunal to establish all facts necessary to resolve the case. 22 This principle is also reflected in Article 20.1 of the ICC Rules, which states that "[t]he Arbitral Tribunal shall proceed within as short a time as possible to establish the facts of the case by all appropriate means."

Moreover, potential evidentiary objections based on the substantive law will generally only be taken into consideration if they are based on the substantive law at the seat. 23 However, as Samir Saleh notes, the absence of rules of admissibility in arbitration introduces "a strong subjective element … when the determination of admissibility is no longer dependent on a provision of law. In the absence of the legal rule, and inasmuch as the arbitrator has the status of a lawyer who tries to avoid arbitrary decisions, the latter is often influenced by his legal background. … As a result of a strong subjective element, the well-known clash of cultures may be increased." 24[Page435:]

ii. Need to address privilege at the stage of admissibility

An analysis of the various types of Article 9 objections also underscores the need to consider questions of evidentiary privilege in a systematic matter ex ante. The Article 9 objections in the IBA Rules may be divided into three categories according to the type of concerns underlying the objection. Of these various grounds for objection, only privilege and confidentiality require legal analysis that, ideally, should be decided or agreed upon in advance.

Pragmatic concerns. Article 9 objections on the grounds of undue burden (Article 9.2(c) of the IBA Rules) and loss or destruction of documents (Article 9.2(d) of the IBA Rules) relate to questions of fact overlaid with fairness issues that can properly be decided by the arbitral tribunal in its discretion. These are fact-specific, pragmatic concerns regarding the possibility of and burden associated with the production of particular evidence that can be dealt with adequately when the objection arises.

Weight concerns. The Article 9 ground for objection due to lack of sufficient relevance or materiality (Article 9.2(a) of the IBA Rules) goes directly to concerns regarding the weight of the evidence. The assessment of relevance and materiality requires no special agreement, and the analysis can only be undertaken on the basis of the knowledge of the case at the time. Evidence that is admitted with insufficient relevance or weight may have caused the party unnecessary effort but can be safely ignored by the arbitral tribunal in its assessment.

Admissibility concerns. By contrast, issues of legal privilege (Article 9.2(b) of the IBA Rules) and commercial or technical confidentiality (Article 9.2(e) of the IBA Rules) must be addressed by the arbitral tribunal in terms of the applicable facts and law at the outset of proceedings, before the evidence has been produced, in order to avoid waiver of the privilege.

As privilege is a threshold admissibility issue, it is not possible to postpone the issue until the stage of the weighing of the evidence, as would be the case with objections based upon other evidentiary questions such as hearsay or authenticity of documents. Simply ignoring all rules of privilege is not an option because it could ultimately undermine the recognition or enforcement of the award on the grounds of public policy. 25[Page436:]

In fact, the need to address the question of the applicable rules of privilege ideally arises long before evidence is objected to in arbitral proceedings. The proposal of defining the applicable privilege in the arbitration clause has been raised frequently in the literature and has been dismissed as impractical as many times. The only real solution to the predictability problem is the formulation of default rules of privilege around which an international consensus could crystallize.

3. Options for approaching evidentiary privilege in international arbitration

The two main options for addressing evidentiary privilege ex ante are by means of (a) a choice-of-law rule for legal impediment and privilege and (b) default rules of privilege. Numerous variants are conceivable within each of these options, both from a rule-making perspective and in the context of a particular arbitration.

As of the writing of this article, it is not known whether the IBA Rules Subcommittee will propose additional guidance on privilege in the IBA Rules or what form that guidance might take. However, no matter what the result, these options will continue to be of practical relevance. If the IBA Rules contain additional guidance on a choice-of-law approach, an arbitrator may determine nevertheless that default rules are sensible in a particular situation or parties may decide that they wish to create certainty at the time of contracting. If the IBA rules contain default rules, these may not be appropriate for certain arbitrations, for example because the parties in a particular arbitration come from two civil-law jurisdictions without extensive discovery or privilege rules, such as France and Germany. Similarly, if the parties come from two common-law jurisdictions such as England and the United States, such rules might accord privilege where the parties had not expected it (e.g. duty of secrecy) and, at the same time, restrict the privilege where both had anticipated its coverage.

a. Disadvantages of the choice-of-law method

The choice-of-law method has been commented upon extensively in the literature. 26 Among the factors found relevant to determining the closest connection are: the jurisdiction of admission of affected attorneys, the jurisdiction of the party asserting the privilege and the jurisdiction with the closest connection to the communication27 or some combination thereof. As a secondary, corrective step, the two options are either a least-favoured-nations rule or, more likely for fairness reasons, a most-favoured-nations ratchet-up. [Page437:]

However, the choice-of-law approach has been rightly criticized as having two major drawbacks. First, it is inherently unpredictable. Even if parties knew ex ante which choice-of-law rules an arbitral tribunal would apply, it could not predict with any certainty how an arbitral tribunal would apply those rules and what result it would reach. 28 Second, this approach can, and regularly does, result in different standards being applicable to various parties or communications, which, if not corrected, can be inherently unfair.

b. Considering default rules of evidentiary privilege

If one is prepared to (i) take the liberty of exercising discretion in the application of the rules of law applicable in a particular jurisdiction in the context of arbitration or (ii) create an MFN rule by assembling a mosaic of rules from different systems, this article proposes that it is but a small step further to acknowledge that default rules created in advance for use in international arbitration can be made applicable in particular cases, with little or no regard to the particular substantive and procedural rules in play in a particular arbitration. As has been shown above, with the possible exception of ethical rules and criminal sanctions applicable in the rare situation in which a counsel itself provides evidence, arbitrators need not fear the applicability of mandatory national law as a possible source of conflict with default rules. Thus, no real limitations exist that prescribe the discretion granted to the arbitral tribunal to find an expedient, just solution.

In the medium to long term, a system of default rules of privilege that could be chosen or anticipated by the parties ex ante would have three major advantages over an adapted choice-of-law approach. First, the results are fully predictable under default rules, whereas a choice-of-law analysis is inherently unpredictable. Even if the nationalities of the parties are known at the outset, the choice-of-law analysis may lead to the law of a third country if that country is most closely connected with the potentially privileged communication. Second, once a standard set of default rules of privilege has been established, there will be significant efficiency and economy reasons for simply choosing it over conducting an individualized choice-of-law analysis in each case. Third, only a system of default rules that can be known ex ante will enable parties to structure their commercial relations on that system and reliably and predictably reap the benefits those sets of rules were designed to achieve (protection of attorney-client relationship, protection of secrets, encouragement of settlement negotiations, etc.). Party expectations and the protections available to parties are two sides of the same coin. The higher the protections available, the firmer the expectation will become as parties structure their legal and commercial affairs in accordance with such protections. [Page438:]

Less attention has been granted thus far to concrete possibilities for transnational or default rules. These default rules could take the form of guidelines or protocols, or a procedural order or orders that contain generic or harmonized rules of privilege that may be used again or tailored to specific contexts. In crafting such default rules, one could take various possible starting points, including a client-held attorney-client privilege (as in the United States), a counsel-held duty of secrecy (as in France, Germany and Switzerland) or a cumulative approach. If part of a protocol or guidelines, default rules could be made generally applicable or take the form of an "opt-in". 29 The default rule could also include an exception for a showing of exceptional need, if it is considered necessary. These and other options will be discussed in the following subsections.

i. Proposal for a cumulative approach to default rules

Nothing in the foregoing has suggested that the attorney-client privilege and the duty of confidentiality or secrecy, in their various iterations, are mutually exclusive. For this reason, as well as for fairness reasons, the simplest approach in constructing default rules of privilege is to combine the various rules in a cumulative manner. Thus, a document could be considered privileged and excluded from evidence if it satisfied either (i) the criteria of the attorney-client privilege or (ii) the duty of confidentiality or secrecy. This approach, similar to the MFN correction to the choice-of-law method, is essentially a ratcheting up of the level of protection available.

The key differences between this approach and the conflict-of-laws approach are as follows.

First, the contours of the rule are, at least for the arbitrator, predictable and do not require the time and effort involved in a conflict-of-laws analysis. Each party essentially receives the level of protection available in their home country or protection that is greater in scope. An arbitrator might develop three versions of default rules to be included in procedural orders. One could be used between parties from civil-law countries, another between parties from common-law countries and a third could be used in arbitrations with parties from both civil-law and common-law countries. Ultimately, international best practices might tend to favour one of these models.

Second, the default rules chosen should not be the choice of a national system of privilege, with all of its intricacies, but rather something new and sui generis. The question should not be whether a document would be privileged under the law of a particular country, which in turn would require extensive analysis of the current state of the law in that jurisdiction. Rather, [Page439:] in keeping with the nature of arbitration as an efficient system of dispute resolution, many of the finer points of national privilege laws can and should be stripped away for use in international arbitration.

ii. Approaches to the problem of waiver

Waiver of privilege is another aspect in which national systems of privilege differ. In most common-law countries, the attorney-client privilege is held by the client, meaning that it is the client who can waive it. By contrast, the duty of secrecy or confidentiality is held by the attorney. In some countries, the attorney may waive it, in others, the client. Yet other countries may deem the privilege to be held by both attorney and client and, hence, to be waivable by both.

How is this issue best addressed in default rules of privilege? On the one hand, waiver is a complex topic that is perhaps best avoided to the extent possible. One option is to retain whatever differences may exist in the sense of the cumulative approach discussed above. The other option would be to say generally that the privilege may be waived and that issues of waiver will be decided by the arbitral tribunal in its discretion.

On the other hand, the idea of waiver could also be utilized in implementing default rules. Javier Rubenstein and Britton Guerrina explored the idea that parties may be deemed to have waived national rules of privilege by virtue of selecting a framework of rules for their arbitration that does not explicitly recognize the privilege. 30 Taking this thought one step further, a default set of privilege rules could incorporate a provision whereby parties expressly waive the privilege to the extent not covered by the agreed rules.

iii. Approaches to the problem of in-house counsel

Another difficult issue is the question of whether or not the privilege should extend to in-house counsel. This question is complicated by the fact that the law in this area is still developing, in particular in Europe. When application of the privilege is made dependent on bar membership, the picture is further complicated by the fact that various jurisdictions have higher bar membership rates than others. In certain countries, for example Japan and Korea, comparatively few attorneys pass the bar and much of the legal work is performed by unadmitted legal professionals. In other jurisdictions, attorneys may no longer remain members of the bar if they go in-house, as they are then no longer considered to be "independent". [Page440:]

In conceiving a default rule or rules, care must be taken not to increase the expectations of one party without increasing the expectations of the other. This would be the case, for instance, if privilege were extended to all admitted in-house counsels in a case involving parties from Switzerland, where the duty of secrecy does not extend to in-house counsel, and Japan, where this is also the case, but where many in-house legal personnel will not be admitted either and hence would not benefit from the new rule. However, the situation would be different if both nationalities benefited from the chosen default rule, as in a case involving parties from Japan and Korea or from France and Switzerland. The value of certainty and ease of decision making may weigh in favour of a simple blanket rule. On the other hand, it may be easier to determine the identities of the various counsels and their status and function on a case-by-case basis.

iv. Exceptions to the attorney-client privilege

Finally, the drafter of default rules of privilege must consider whether the rule ought to have an exception for the showing of cause. In the United States, the attorney-client privilege, while narrowly circumscribed and restrictively applied, is not balanced against another party's need for the document. This is in contrast to the work-product doctrine, which can be overcome upon a showing of special need.

In international arbitration, flexibility would speak in favour of a rule with a clear exception at the discretion of the arbitral tribunal. In international arbitration, the parties' expectations with respect to preservation of the privilege must ultimately be balanced against the responsibility of the arbitral tribunal to marshal the evidence necessary to decide the case. Additionally, exceptions are necessary, as they are found in most national systems, to prevent the invocation of privilege to cover up suspected money laundering or the rendering of legal advice to accomplish a fraud or other illegal purpose.

v. Other evidentiary privileges

Although the attorney-client privilege probably presents the most difficult challenges, it should not be forgotten that there are a number of other evidentiary privileges.

Most of these evidentiary privileges will not be asserted with any regularity in international commercial arbitration. These include, for example, the communications between spouses, which are protected in many countries, [Page441:] and the communications between religious leaders and persons seeking advice (priest-penitent privilege). Such privileges can probably be safely omitted from default rules of privilege for use in international commercial arbitration and dealt with on an ad hoc basis should they arise.

However, two other privileges will find frequent application and should be included in default rules of privilege.

First, the settlement or "without prejudice" privilege should be given effect, in particular as, of all the evidentiary privileges, it has been found to rest on transnational principles. 31 The default rule should make provision for waiver in case all parties to the settlement negotiations agree to the disclosure. Given the international setting, it is not advisable to insist on particular "magic words". Rather, it is preferable to leave this privilege broadly worded and tied to the purpose of the communications, rather than de facto requiring a "without prejudice" or similar legend.

Second, the litigation privilege (as it is known, for example, in England) or work-product doctrine (as it is known, for example, in the United States) is likely more broadly accepted internationally than the attorney-client privilege. This is because the communications or documents in question will often relate directly to the dispute at hand rather than to legal advice rendered in other contexts, the legitimacy of which may be questioned. However, the litigation privilege can complicate matters and potentially broaden the scope of privilege, as it may not always be tied to the involvement of an attorney under national rules. For arbitrations in which arbitrators determine that the scope of privilege should be more limited, the work-product doctrine can probably be omitted on the theory that most documents potentially covered by it will be covered by the attorney-client privilege. For arbitrations involving parties familiar with the litigation privilege or work-product doctrine, it could be safely included, provided there is a clear exception for a showing of need.

4. Conclusion

If the new version of the IBA Rules contains default rules of privilege, the question will still remain as to whether parties and arbitrators should adopt them in a particular case. Similarly, if default rules are not adopted in the IBA Rules themselves, parties and arbitrators may consider whether they should be formulated and adopted in procedural orders in particular cases. [Page442:]

Privilege issues are likely to arise with increasing regularity in international arbitration in the coming years. Some have speculated that any change in the IBA Rules relating to privilege would raise awareness of the issue, possibly encouraging parties to raise privilege objections more regularly. Additionally, as information in electronic form becomes a more regular aspect of international arbitration, instances of inadvertently disclosed privileged documents will arise more frequently. This problem may be addressed with so-called "claw-back" agreements for the return of privileged materials inadvertently disclosed in an electronic production. However, such agreements also presuppose a certain consensus on the extent of the applicable legal privilege. Under these circumstances, arbitrators will find it increasingly necessary to develop a clear approach to the question of privilege that may then be adapted to the situation in a particular arbitration.

Given the current lack of clarity on the topic, it is to be hoped that the IBA Rules will provide additional guidance on the determination of the applicable rules of legal impediment or privilege. Whether or not the IBA Rules are expanded, arbitrators should become more proactive in addressing the issue, either through choice-of-law approaches or through default privilege rules. Any of these options will, over time, contribute to a growing international consensus as to best practices. Increasing familiarity with privilege determinations will assist arbitral tribunals in making such determinations more predictably, efficiently and economically. Ultimately, standard default rules of legal impediment and privilege may become established enough that parties will either agree on them up-front or have a high expectation of being able to rely on such privileges even in the absence of express agreement. Only the ex ante assurance that privilege will be recognized in a possible arbitration will enable parties to structure their commercial affairs to take account of that privilege, thus achieving the policy objectives underlying the various rules of privilege.

1
In early 2008, then Co-Chairs of the IBA Arbitration Committee Sally Harpole and Pierre Bienvenu had formed the Subcommittee and appointed Richard H. Kreindler to serve as Chairman. The membership of the subcommittee included 17 experienced practitioners from around the world, and the subcommittee was advised by members of the working group responsible for proposing the 1999 version of the IBA Rules. As at the date of publication, the membership of the IBA Rules Subcommittee included: Richard H. Kreindler (Chairman), David Arias, Mark C. Baker, Antonias Dimolitsa, Paul Friedland, Nicolás Gamoba, Peter Heckel, Stephen Jagusch, Xiang Ji, Kap-You (Kevin) Kim, Toby T. Landau, Alexis Mourre, David W. Rivkin, Georg von Segesser, Essam Al Tamimi, Hiroyuki Tezuka and Ariel Ye.

2
All references to the IBA Rules are to the 1999 version, the 'IBA Rules on the Taking of Evidence in International Commercial Arbitration', adopted by a resolution of the IBA Council on June 1, 1999, which are reprinted in this volume and are currently available on the IBA website at: http://www.ibanet.org.

3
Bernard Hanotiau, 'Document production in international arbitration: A tentative definition of "best practices"', ICC Bulletin 2006 special supplement: Document production in international arbitration (2006) p. 113 at p. 117.

4
IBA Working Party, 'Commentary on the new IBA Rules of Evidence in International Commercial Arbitration', 2 Business Law International (2000) p. 16 at p. 36.

5
Javier H. Rubinstein and Britton B. Guerrina, 'The attorney-client privilege and international arbitration', Journal of International Arbitration (2001) p. 587 at pp. 598-99 (advocating a MFN correction to the choice-of-law approach).

6
Various survey resources are available to arbitrators when assessing the differences between the rules applicable in various jurisdictions. See, e.g., Lex Mundi, 'In-house counsel and the attorney-client privilege', available at: https://www.lexmundi.com/images/lexmundi/PDF/ AttyClient/2007_Atty_Client_Update/Attorney_Client_update8.07.pdf (last updated August 2007) (surveying the rules of privilege and confidentiality in over 85 countries as well as the European Union and US states with special emphasis on the applicability of such rules to in-house counsel); Markus Koehnen, Marc Russenberger and Eric Cowling, Privilege and confidentiality: An international handbook (IBA 2006) (surveying the laws of privilege and confidentiality in 18 jurisdictions). The references to national laws in this article are largely drawn from the Lex Mundi survey.

7
Hans Smit, 'Roles of the arbitral tribunal in civil law and common law systems with respect to presentation of evidence', 7 ICCA Congress Series (1996) p. 161 at p. 168.

8
IBA Working Party, 'Commentary on the new IBA Rules of Evidence in International Commercial Arbitration', 2 Business Law International (2000) p. 16 at p. 19.

9
Klaus-Peter Berger, 'Evidentiary privileges: Best practice standards vs./and arbitral discretion, Best practices in international arbitration (2006) p. 19 at p. 21.

10
Pierre Heitzmann, 'Confidentiality and privileges in cross-border legal practice: The need for a global standard', 26 ASA Bulletin (2008) p. 205 at p. 207.

11
See Klaus-Peter Berger, 'Evidentiary privileges: Best practice standards vs./and arbitral discretion, Best practices in international arbitration (2006) p. 19 at p. 25 (citing to Henssler and Hellwig).

12
See, e.g., Rule 8.5(a) of the ABA Model Rules of Professional Conduct, which were adopted recently in the State of New York (as of April 1, 2009) ("A lawyer admitted to practice in this state is subject to the disciplinary authority of this state, regardless of where the lawyer's conduct occurs. A lawyer may be subject to the disciplinary authority of both this state and another jurisdiction where the lawyer is admitted for the same conduct.").

13
See Lex Mundi, 'In-house counsel and the attorney-client privilege', available at: https://www.lexmundi.com/images/lexmundi/PDF/AttyClient/ 2007_Atty_Client_Update/Attorney_Client_update8.07.pdf (last updated August 2007).

14
Natalie Voser, 'Harmonization by promulgating rules of best international practice in international arbitration', 3 SchiedVZ (2005) p. 113 at p. 118 (naming privilege as one specific procedural issue in need of harmonization).

15
Michelle Sindler and Tina Wüstemann, 'Privilege across borders in arbitration: Multi-jurisdictional nightmare or a storm in a teacup?', ASA Bulletin (2005) (arguing that privilege is not a proper subject for harmonization); Pierre Heitzmann, 'Confidentiality and privileges in cross-border legal practice: The need for a global standard', 26 ASA Bulletin (2008) p. 205 (offering eight guiding principles for arbitrators in determining and observing the applicable privilege as well as case studies applying them).

16
Klaus-Peter Berger, 'Evidentiary privileges: Best practice standards vs./and arbitral discretion, Best practices in international arbitration (2006) p. 19 at p. 38 (finding that consensus is emerging as to key aspects of addressing privilege in international arbitration and that there is no need for further harmonization).

17
Javier H. Rubinstein and Britton B. Guerrina, 'The attorney-client privilege and international arbitration', Journal of International Arbitration (2001) p. 587 (advocating an MFN-choice-of-law approach); Olaf Meyer, 'Time to take a closer look: Privilege in international arbitration', Journal of International Arbitration (2007) p. 365 (distinguishing between the goals of truth and justice, the formal truth, the more probable version of the truth and the truth that is capable of being found within a reasonable time and at reasonable cost).

18
Olaf Meyer, 'Time to take a closer look: Privilege in international arbitration', Journal of International Arbitration (2007) p. 365 at pp. 370-371 (distinguishing between the goals of truth and justice, the formal truth, the more probable version of the truth and the truth that is capable of being found within a reasonable time and at reasonable cost).

19
William W. Park, 'The 2002 Freshfields Lecture-Arbitration's protean nature: The value of rules and the risks of discretion', 19 Arbitration International (2003) p. 279 at p. 283.

20
Samir A. Saleh, 'Reflections on admissibility of evidence: Interrelation between domestic law and international arbitration', Arbitration International (1999) p. 141 at p. 155.

21
Julian D.M. Lew, Loukas A. Mistelis and Stefan M. Kröll, Comparative international commercial arbitration (2003) pp. 565-566 at para. 22-45 and p. 570 at paras. 22-64.

22
Alan Redfern and Martin Hunter with Nigel Blackaby and Constantine Partasides, Law and practice of international commercial arbitration (2004) p. 296 at para. 6-65.

23
Emannuel Gaillard and John Savage (eds.), Fouchard Gaillard Goldman on international commercial arbitration (1999) p. 694 at para. 1266.

24
Samir A. Saleh, 'Reflections on admissibility of evidence: Interrelation between domestic law and international arbitration', Arbitration International (1999) p. 141 at p. 156.

25
George Burn and Zara Skelton, 'The problem with legal privilege in international arbitration', 72 Arbitration (2006) p. 124 at p. 128.

26
See, e.g., Klaus-Peter Berger, 'Evidentiary privileges: Best practice standards vs./and arbitral discretion, Best practices in international arbitration (2006) p. 19; Pierre Heitzmann, 'Confidentiality and privileges in cross-border legal practice: The need for a global standard', 26 ASA Bulletin (2008) p. 205; and Javier H. Rubinstein and Britton B. Guerrina, 'The attorney-client privilege and international arbitration', Journal of International Arbitration (2001) p. 587.

27
Klaus-Peter Berger, 'Evidentiary privileges: Best practice standards vs./and arbitral discretion, Best Practices in International Arbitration (2006) p. 19 at p. 38 (providing four guidelines for the exercise of the tribunal's discretion regarding privilege).

28
Javier H. Rubinstein and Britton B. Guerrina, 'The attorney-client privilege and international arbitration', Journal of International Arbitration (2001) p. 587.

29
See William W. Park, 'The 2002 Freshfields Lecture-Arbitration's protean nature: The value of rules and the risks of discretion', 19 Arbitration International (2003) p. 279 at p. 290 (discussing the general possibility of supplementary "opt-in" rules in cases in which "'procedure heavy' provisions seem too radical").

30
Javier H. Rubinstein and Britton B. Guerrina, 'The attorney-client privilege and international arbitration', Journal of International Arbitration (2001) p. 587 at pp. 591-592.

31
Klaus-Peter Berger, 'Evidentiary privileges: Best practice standards vs./and arbitral discretion, Best Practices in International Arbitration (2006) p. 19 at p. 32.