There is no doubt that the content of contemporary customary international law has been shaped by the numerous BITs entered into by States as acknowledged by tribunals in cases such as CME Czech Republic B.V. v. Czech Republic (UNCITRAL, Award, 14 March 2003, para. 498) and Mondev v. United States (ICSID, Award, 11 October 2002, para. 125). However, while it is certainly true that BITs will influence the development of customary international law, it is quite another thing to simply say that BITs now represent the new custom in international investment law. The better view is that custom in the field of international investment law does not correspond to the total sum of more than 2,500 BITs. In this respect, the approach was adopted by the Pope & Talbot Tribunal must be rejected (UNCITRAL, Award, 31 May 2002, para. 62: “applying the ordinary rules for determining the content of custom in international law, one must conclude that the practice of States is now represented by those treaties [i.e. BITs].”).
The main weakness of the proposition equalling BITs to new custom is its basic failure to meet the definition of customary international law. Custom has two constitutive elements: a “constant and uniform” (but not necessarily unanimous) practice of States in their international relations and the belief that such practice is required by law (opinio juris).
State practice. In theory, BITs can serve as evidence of the element of State practice required to establish the existence of a rule of customary international law (see, Camuzzi v. Argentina, ICSID, Award, 11 May 2005, para. 144). At the same time, as explained by the recent work of the International Law Association (ILA) on customary international law “[t]here is no presumption that a succession of similar treaty provisions gives rise to a new customary rule with the same content.” (p. 47). The undeniable reality is that BITs are very diverse in their content and scope. They are certainly not consistent enough to constitute the basis for any rule of customary international law. For instance, the inconsistency of State practice is undeniable with respect to the type and scope of legal protection offered to different types of shareholders (minority, indirect) as well as for holding (or “shell”) companies (see: T. Gazzini, “The Role of Customary International Law in the Protection of Foreign Investment”, 8(5) Journal of World Investment & Trade, 2007, at p. 707-710). This is also the position held by many authors in doctrine who have undertaken the analysis of BITs to determine whether specific substantive rights contained in these treaties represent custom (see: M. Sornarajah, The International Law on Foreign Investment, 2nd ed., Cambridge U.P., 2004, p. 206, 220-227, 436, 441-443 ; B. Kishoiyian, “The Utility of Bilateral Investment Treaties in the Formulation of Customary International Law”, 14(2) Northwestern Journal of International Law & Business, 1994, at p. 343-346, 363-373; A. Al Faruque, “Creating Customary International Law Through Bilateral Investment Treaties: A Critical Appraisal”, 44 Indian J. Int’l L., 2004, at p. 304-305, 356-363).
Opinio juris. As explained by the Tribunal in UPS v. Canada, there is no evidence of any “general sense of obligation” by States entering into BITs (UNCITRAL, Award, 22 November 2002, para. 97). In fact, the evidence suggests that the decision of States to enter into BITs is solely based on their (perceived) economic interest. Clearly, developing States sign BITs to attract foreign investments. BITs are the result of trade-offs and mutual concessions between States. Their content depends on the political and economic bargaining power of each party to the negotiations. BITs are the product of a compromise between conflicting interests; they are not entered into by States based on any perceived legal obligation.
It is noteworthy that the proposition that customary law is coterminous with BITs has been explicitly rejected by States. This is clearly the case in the context of NAFTA arbitration where unambiguous statements to that effect have been made by Mexico (submission filed in the context of Loewen v. United States, para. 33 & 39), by the United States (submission filed in: Glamis v. United States, p. 142 et seq.) and by Canada (submissions filed in: Chemtura v. Canada, at para. 269-273, and in: Loewen v. United States, at para. 18).
In sum, the content of custom and the thousands of BITs are simply not the same. Does this mean, as recently concluded by the ADM v. Mexico Tribunal (ICSID, Award, 21 November 2007, para. 117) that BITs only create lex specialis rules solely applicable between the countries which are party to these BITs? conclusion is not entirely satisfactory either, as it wrongly excludes the role that these treaties might play in the development of custom. Thus, these numerous BITs will necessarily influence customary international law.
The impact of BITs on customary international law is twofold. First, some of the standards of protection systematically contained in BITs will certainly contribute to the consolidation of already existing rules of custom in international investment law (see Generation Ukraine v. Ukraine, ICSID, Award, 16 September 2003, para. 11.3). Second, the common features to investment protection resulting from BITs will also contribute to the crystallisation of new rules of customary international law in the future. By their very nature, customary rules evolve over time. The repetitive enunciation of some of the standards of protection existing under BITs may be the starting point of State practice which will eventually become custom. A treaty provision may, indeed, provide the impulse for the formation of new custom.
University of Ottawa (Civil Law Section)