It a well-known fact that investment treaties are short on detail. Similar to nature, law abhors a vacuum, and disputes under these treaties have to be solved somehow. Therefore, arbitrators are often left with significant interpretative space left open by open-ended treaty norms. In the context of substantive protection, this tendency to “fill out” vague norms is much-studied (and criticized): the development of the jurisprudence concerning the scope of the fair and equitable treatment standard is one obvious example that springs to mind. The same gap-filling tendency applies also to procedural matters, however. Arbitrators are regularly exercising discretion in determining procedural details in treaty-based disputes. Assuming that the parties have not come to a specific agreement – which is rarely the case in a scenario where the basic procedural frames are determined by a treaty short on details and where there is typically no prior relationship between the disputing parties – tribunals are entrusted with considerable discretion to, for example, determine the place of arbitration, the rules for document production or the scope for ordering interim measures.
One of the clearest examples of this procedural discretion is the apportioning of the costs of the arbitration. The fundamental issue of who pays what is often not agreed upon by the parties, neither directly (i.e. in the treaty or a post-dispute agreement) nor indirectly (i.e. in the applicable arbitration rules or in the lex loci arbitri, both of which, as will be seen below, tend to be rather open-ended).