The exchange of discovery, generally, let alone e-discovery, specifically, is uncommon in investor-state arbitration (arbitration between a foreign investor and the host sovereign state of the investment, generally conducted pursuant to an investment treaty). Indeed, practitioners prefer to employ the term “document production” to describe the concept of discovery in these arbitrations.
The main reason why the concept of discovery is uncommon in these investor-state arbitrations is that, typically, the parties come from various legal backgrounds, including civil law traditions where document production is extremely limited. In investor-state disputes, each party is primarily responsible for establishing the facts of its case, contrary to the common law tradition where parties may request considerable production of documents.
However, complex issues of document production do sometimes occur in investor-state disputes. Parties and arbitrators faced with such issues can rely on rules from various sources. Moreover, document production in an investor-state dispute requires a specific approach, owing to the fact that a sovereign state is always a party in the proceeding.
LEGAL GROUNDS FOR REQUESTING DOCUMENTS
The Washington Convention of 1965 created the International Centre for Settlement of Investment Disputes (ICSID), which administers most investor-state disputes. At the outset, it must be noted that neither the ICSID Convention nor the ICSID Arbitration Rules address discovery. Similarly, the UNCITRAL Arbitration Rules, which are generally employed in ad hoc investor-state disputes, are also silent on the subject.
There are, however, rules on document production. Article 34(2)(a) of the ICSID Arbitration Rules provides that “the Tribunal may, if it deems it necessary at any stage of the proceeding, … call upon the parties to produce documents, witnesses and experts.” Furthermore, Article 24(3) of the UNCITRAL Rules mentions that “at any time during the arbitral proceedings the arbitral tribunal may require the parties to produce documents.”
Despite these provisions, many questions remain unanswered when arbitral tribunals have to deal with requests for the production of documents in investor-state disputes. These requests happen frequently: in 2009, for example, 12 of the 21 publicly available awards granted in the course of investor-state disputes (excluding decisions on annulment) included a document production procedure.
Generally, the parties agree to rely on the International Bar Association Rules on the Taking of Evidence in International Arbitration (the IBA Rules), as a non-binding reference, as early as the first session with the arbitral tribunal, which takes place mandatorily within two months of the constitution of the tribunal. The parties and the arbitral tribunal then refer to the IBA Rules in deciding on these issues.1 The IBA Rules are thus considered to be acceptable for both common law and civil law parties and their representatives, in that they represent a middle ground between the two legal cultures. Not surprisingly, in all relevant, published investor-state awards, the arbitral tribunals ordered narrow and specific document production as promoted in the IBA Rules.
SPECIFIC FEATURES OF DOCUMENT PRODUCTION IN INVESTOR-STATE DISPUTES
An arbitration procedure involving a state is not a typical day-to-day commercial dispute. This is particularly true of investor-state arbitration as it concerns a state’s acts that are performed in the exercise of its sovereign rights over territory (such as expropriation, enactment or revocation of laws, statutes or regulations) or sometimes in the context of local political or economical crises. As such, these proceedings may relate to particularly sensitive issues in the host state, and this may, in certain cases, lead to obstacles at the document production stage.
One way that states seek to avoid producing requested documents is to rely on the exception of classified information. The North American Free Trade Agreement (NAFTA), Chapter 11 of which is the basis for many investor-state disputes, contains a specific provision on this matter. Indeed, Article 2102 provides that “nothing in this Agreement shall be construed to require any Party to furnish or allow access to any information the disclosure of which it determines to be contrary to its essential security interests.” Some bilateral investment treaties (BITs) also contain similar terms. The Canada-Jordan BIT of 2009, for example, provides that “the Tribunal shall not require a Party to furnish or allow access to information the disclosure of which would impede law enforcement or would be contrary to the Party’s law protecting Cabinet confidences, personal privacy or the financial affairs and accounts of individual customers of financial institutions, or which it determines to be contrary to its essential security.”
Arbitral case law shows that states sometimes try, with some success, to extend the scope of the state secrets privilege by relying on provisions contained in their own domestic law. The tribunals’ reactions on the subject are disparate. In Pope & Talbot v. Canada (a NAFTA-based UNCITRAL case), the state relied on the Canadian notion of “cabinet confidence” to object to the production of a range of documents. The NAFTA tribunal judged the objection to be valid where state secrets were concerned, but it rejected the application of Canadian law, which provided for a broader scope of privilege, by deeming it inapplicable to an international arbitral tribunal. Yet in Glamis Gold v. USA (also a NAFTA-based UNCITRAL case), the tribunal upheld the USA’s objection related to “deliberative process privilege” with regard to various pre-decisional and deliberative documents comprising part of the process by which governmental decisions and policies were formulated.
In Biwater Gauff v. Tanzania (a BIT-based ICSID case), the state objected to the production of various documents on the basis of the Tanzanian principle of “public interest immunity.” This principle allegedly applied to unpublished official records and communications received by a public officer, the disclosure of which would be prejudicial to public interest. The tribunal rejected the objection, noting that the objection was based on Tanzanian law, which was not applicable to an international arbitral tribunal.
Another complication regarding investor-state arbitration is that the investor is often a company set up specifically for the requirements of the investment and may otherwise be inactive (e.g., a shell company). A document production order against such a company will bear little fruit, as the substantial documentary evidence will be held by those actually controlling that company. As a third party to the arbitration, this person or entity is beyond the arbitral tribunal’s jurisdiction and cannot be subjected to an order to produce the requested documents.
Within their own borders, states are in a stronger position to obtain documents despite a pending international arbitration. In this context, certain states have been known to resort to their own legal systems to obtain documents, notably those belonging to the investor. By employing “help-yourself” discovery tactics, a state can effectively bypass the arbitral tribunal and its powers to decide on the production of documents. For example, a state may seize the investor’s documents in the context of criminal proceedings, whether initiated on valid grounds or as a means of pressuring or retaliating against the investor. In this case, the investor may find itself in the very awkward position of having to request, before the arbitral tribunal, the production of its own documents, seized and sealed by the police or related authorities.
Finally, investors should be cautious as to what they can expect to receive in terms of document production, particularly in the form of electronic documents. Very few states have centralized or unified information technology (IT) systems. Thus, there is a strong possibility that IT policies regarding storage and destruction of information differ from department to department as well as from nation to nation. Moreover, multiple servers and/or databases may need to be searched for potentially relevant documents. As a result, electronic searches may be considered overly burdensome, and thus may be rejected by arbitral tribunals.