Legal privilege issues: at the mercy of the arbitral tribunal

Thomas Stouten
Houthoff, Rotterdam
t.stouten@houthoff.com

Denise Jansen
Houthoff, Rotterdam
​​​​​​​denise.jansen@houthoff.com

Most jurisdictions recognise the principle that communications between lawyers and their clients must be respected. This principle is often referred to as ‘legal privilege’.[1] In the context of national court proceedings, it will usually be clear what such legal privilege entails (exactly) according to the applicable national legal privilege rules. However, things become complicated in the context of international commercial arbitration. This is due to the fact that a single international commercial arbitration will generally involve lawyers from different jurisdictions in concert with the different conceptions and interpretations of legal privilege across the world.

This article will elucidate on the methods arbitral tribunals can employ to address issues related to legal privilege. At the outset, we will briefly discuss the possibilities for parties to agree upon the applicable national legal privilege rules. Then, we will deal with (the absence of) notions of legal privilege under the major arbitration rules. By extension, this contribution touches upon the guidance, albeit limited, provided by the (recently revised) IBA Rules on the Taking of Evidence in International Commercial Arbitration (the IBA Rules). It subsequently explores how arbitral tribunals should determine the applicable national legal privilege rules in a case at hand, including alternative approaches such as (1) the application of the 'closest connection' test and (2) the 'most' and 'least' favourable legal privilege rules approach. In its conclusion, this article considers how the predictability of legal privilege may be improved in practice.

Agreeing on legal privilege rules when concluding the arbitration agreement

Party autonomy is a cornerstone of international arbitration. Accordingly, like other matters in arbitration, parties may agree upon the applicable national legal privilege rules governing their lawyer–client communications. Since dispute resolution mechanisms however are usually included in the contract at the last minute (hence their reference as so-called ‘midnight clauses’) or as an afterthought, there is rarely the time or attention given to discuss and agree on issues such as legal privilege in a possible future dispute. When a contract does not expressly include a provision on legal privilege, the matter may however still be covered by the arbitration rules chosen by the parties, insofar as those rules address legal privilege.

(Institutional) arbitration rules on legal privilege

As briefly noted above, there is no uniformity on legal privilege rules across the world. Differences may be found when dealing with questions such as whether the client or the lawyer holds the privilege, whether work product falls within the ambit of legal privilege, whether the privilege is applicable to in-house counsel (or not) and whether or to what extent exceptions to legal privilege may apply. Given these sorts of discrepancies, parties’ expectations can easily diverge from the national privilege rules that an arbitral tribunal ultimately applies. Consequently, one might expect international arbitration rules to provide further guidance. In practice, however, leading institutional arbitration rules are largely or completely silent on the issue of legal privilege.[2] Most, if not all, institutional rules are designed to be applied in multiple jurisdictions, and legal privilege may be further complicated depending on the number of international aspects of a particular case. Consequently, it requires a case-by-case analysis. 

There are at least two sets of (institutional) arbitration rules that explicitly address legal privilege to a certain extent: the 2018 International Institute for Conflict Prevention and Resolution Rules for Non-Administered Arbitration (CPR Rules) and the International Centre for Dispute Resolution (ICDR) Rules. Notably, while both sets of rules are US-centric, they take differing approaches to legal privilege. Whereas the CPR Rules give broad discretion to the arbitral tribunal to deal with issues surrounding legal privilege,[3] the ICDR Rules stipulate that the arbitral tribunal should (to the extent possible) give ‘preference to the rule that provides the highest level of protection’.[4] This latter approach is referred to as the 'most favourable' – it could also be referred to as the most cautious – approach to legal privilege.  It will be covered when discussing the approaches available to arbitral tribunals in determining the applicable legal privilege rules. 

Guidance provided in the IBA Rules

Since most (institutional) arbitration rules are silent on legal privilege, soft law instruments such as the IBA Rules fill the gap and provide further guidance. Article 9.2 of the 1999 version of the IBA Rules addressed legal privilege as one of the reasons for excluding a document from evidence or production. The 2010 version of the IBA Rules additionally provide that the arbitral tribunal may take into consideration:

‘(a) any need to protect the confidentiality of a Document created or statement or oral communication made in connection with and for the purpose of providing or obtaining legal privilege;

[…]

(c) the expectations of the Parties and their advisors at the time the legal impediment or privilege is said to have arisen;

(d) any possible waiver of any applicable legal impediment or privilege by virtue of consent, earlier disclosure, affirmative use of the Document, statement, oral communication or advice contained therein, or otherwise; and

(e) the need to maintain fairness and equality as between the Parties, particularly if they are subject to different legal or ethical rules.’ [5]

The 2020 revision of the IBA Rules makes no changes to the abovementioned considerations.[6]

Nonetheless, not even the IBA Rules provide any guidance on how an arbitral tribunal might determine which national legal privilege rules to apply, leaving this discretion to the arbitral tribunal. The commentary of the 1999 IBA Working Party & 2010 IBA Rules of Evidence Review Subcommittee acknowledge this discretion,[7] and the commentary on the revised text of the 2020 IBA Rules again remains the same. In the same vein, Article 3.9 of the IBA Rules (in both the 2010 and 2020 versions) stipulates that the steps to be taken following a request for documents shall be those the arbitral tribunal ‘considers appropriate’.

How to determine the applicable national legal privileges rules (if any)

With a limited legal framework (and in the absence of the parties' explicit agreement), arbitral tribunals are left to determine which national legal privilege rules to apply (if any). When doing so, they have ample room to consider the best approach to be pursued. However, considering that arbitral tribunals have the duty to issue an enforceable award, they are in any event restricted by underlying principles of arbitration that may impact such enforceability, such as the principle of equal treatment of the parties and due process considerations.

When starting an analysis on which national legal privilege rules to apply, one should consider whether legal privilege is procedural or substantive in nature. There is no clear-cut answer to this question. From a Dutch perspective, legal privilege is qualified as a procedural law issue.[8] Other civil law jurisdictions, for example Germany, Belgium and Italy, similarly deem legal privilege a procedural matter, as it also relates to the limitation to disclose evidence – which limitation is considered a matter of procedural law.[9] To the contrary, most common law jurisdictions, for instance Canada, Hong Kong, Thailand, the United States and the United Kingdom, qualify legal privilege as a substantive law issue, as certain types of legal privilege are considered substantive protections. Various commentators have argued that in reality, legal privilege is both procedural and substantive in nature.[10]

Procedural issues in international arbitration are generally governed by the lex arbitri (ie, the law of the place or ‘seat’ of the arbitration). Issues that are substantive in nature will generally be dealt with by applying the law of the underlying contract or the law determined via the rules of private international law (ie, conflicts of law). In the absence of any specific agreement on legal privilege, the question arises whether the parties intended to apply the national legal privilege rules of the lex arbitri or the national legal privilege rules under, for example, the law governing the contract. It is questionable whether the parties would have had any intention to apply either of these laws, in particular considering that other factors, such as the locations where the parties reside or their lawyers' practice, will likely be of greatest import to the parties when deciding on how legal privilege should be dealt with.

This would be reflected in practice in a situation in which the parties have chosen the seat (and therefore the lex arbitri) and/or the law governing the contract only based on its neutrality, despite the parties (and their lawyers) having no real connection thereto. Applying the national legal privilege rules purely based on the location of the seat or under the law governing the contract could be contrary to the parties' expectations.

In order to remedy the shortcomings of the abovementioned approach, a tribunal may turn to two alternative approaches when deciding on the issue of applicable national legal privilege rules. These will be outlined further below.

'Closest connection' approach

One possible approach is to identify the national legal privilege rules pursuant to 'the closest connection' test, pursuant to which an arbitral tribunal looks to establish an objective connection between the privilege claimed and particular national legal privilege rules.[11] The key question here is what facet of legal privilege should be taken into consideration when determining the national legal privilege rules with the closest connection thereto. Myriad factors may be considered when determining such connection, including where a relevant document was created, where the lawyer that created the document is qualified to practice or where (ie, in which jurisdiction) the client resides.[12]​​​​​​​

Approaching legal privilege from each of these factors comes with its own pitfalls. For instance, assuming the lawyer's domicile relevant for the outcome of legal privilege might promote forum shopping when choosing counsel. Although the outcome of possible legal privilege issues will usually not be decisive in this regard, it will nonetheless incentivise parties in their choice of counsel.

Focusing on the 'closest connection' towards the relevant documents bears its own shortcomings. When focusing on this factor, an arbitral tribunal may have to determine the 'closest connection' for each individual document for which legal privilege is claimed. Such an exercise would be extremely time-consuming and would likely obstruct an efficient (and expeditious) flow of the arbitral proceedings. More importantly, focusing on the 'closest connection' towards relevant documents will allow different laws to be applied to different sets of communications or documents in the same arbitral proceedings, further complicating the practical application of legal privilege.

By way of illustration, when a party claims legal privilege in relation to ten separate documents, some of which were created in the Netherlands, while others were created in the US, France and the UK, this would result in four different sets of national legal privilege rules being applied in the same arbitral proceedings, perhaps even to a single document production request. In these circumstances, the 'closest connection' test would create confusion for the parties involved rather than providing legal certainty in the determination of national legal privileges rules. A more harmonised approach providing more certainty is outlined below. 

Applying the 'most' or 'least' favourable legal privilege approach

Arbitral tribunals may also determine applicable national legal privilege rules by choosing the 'most favourable' or the 'least favourable' regime available from the list of national legal privilege rules that may be applicable in a particular arbitral proceeding. Under the 'most favourable' legal privilege approach, the arbitral tribunal will apply the national legal privilege rules providing the widest protection. The 'least favourable' legal privilege approach operates similarly to the 'most favourable' legal privilege approach, but instead applies the rules that offer the lowest standard of protection. In other words, this approach would tend to favour disclosure rather than the shielding of evidence.

An advantage of both approaches is that the parties are treated equally. Both approaches, however, also have their disadvantages which the arbitral tribunal needs to consider carefully when deciding on whether to adopt either approach.

If the 'most favourable' national legal privilege rules approach would be followed, an arbitral tribunal could apply national legal privilege rules that provide for greater protection than the national legal privilege rules that a party (and by extension its counsel) would have assumed to be applicable. In this sense, the 'most favourable' approach might be criticised as overly cautious in its protection of evidence. Another shortcoming is that this approach could hinder the arbitral tribunal's task of establishing all relevant facts and providing a fair opportunity for both parties to present their cases.

The 'least favourable' approach equally has its downsides. When deploying this approach, there is a risk that a party would argue that it is not treated equally or fairly, for instance when it would be entitled to a greater degree of legal privilege protection under its 'home' national legal privilege rules or the rules with the closest connection to its lawyer–client communications. In practice, this approach may lead to a situation in which a party legitimately assumed legal privilege to be applicable to certain documents, yet is nonetheless required to disclose these documents since the arbitral tribunal chose to apply national legal privilege rules that offer a lesser degree of protection.

Moreover, this approach may bear challenges for the parties' lawyers considering their own ethical obligations of confidentiality and/or advice that they provided on the basis of their legitimate belief in the applicability of certain privilege rules. If the privilege rules applied by the arbitral tribunal would be less far-reaching than the local bar's ethical rules, a situation may arise where a lawyer is obliged to disclose documents under the national legal privilege rules applied by the arbitral tribunal, while at the same time mandated to refrain from doing so based under the local rules to which his or her law license is subject.

Towards more predictability on applicable national legal privilege rules

The approaches mentioned above may help the arbitral tribunal to narrow down the possible laws to be applied to legal privilege issues. At the same time, there is still ample room for improvement in terms of predictability and legal certainty. Based on the analysis above, one may conclude that a one-size-fits-all solution is not feasible considering the multi-variable nature of arbitration. Yet, there are several ways that predictability and certainty could nonetheless be enhanced.

A best approach would be for the parties to jointly agree on the applicable privilege rules at the outset of the arbitral proceedings, if they have not already done so in the contract containing their arbitration agreement. In this context, it may also be helpful if arbitral tribunals would distribute a list of potential legal issues that the parties may consider at the outset of the arbitral tribunal, including the issue of legal privilege. Additionally, an arbitral tribunal could request the parties to provide their views on legal privilege briefly, further to which it may educate the parties of any specific (future) issues that could arise due to these differences.

Alternatively, the arbitral tribunal itself could exercise its discretion and inform the parties of its approach to any issues of legal privilege that may arise during the arbitral proceedings. Again, it would be most helpful to address this matter with the parties at the outset of the proceedings, for example during the first procedural call or introduced in a draft procedural order. Such a pragmatic approach would give the parties much needed clarity on the issue well before a document production phase, let alone when document production is at full speed.  

Nevertheless, in the absence of party or tribunal action, most often parties ultimately still remain at the mercy of their arbitral tribunal in their approach to legal privilege. Arbitral institutions might help take away some of this certainty by following the ICDR example and adopting some approach to privilege issues in the rules, irrespective of which one. If legal privilege would be set out in the arbitral institutional rules, the parties would – at least in theory – have access to and the ability to see this approach ahead of time. That may provide the parties as much certainty as they can get without any party agreement on legal privilege.