It has become fashionable in recent years, each time an ICSID annulment decision is released that takes issue with the procedures or reasoning of an ICSID tribunal, for commentators to bemoan the lack of certainty, predictability and finality that this reflects in the ICSID system for adjudicating investment treaty disputes between investors and host States. Some commentators urge a return to greater use of ad hoc UNCITRAL arbitration, or arbitration before institutions other than ICSID, to avoid the perceived vagaries of the ICSID annulment process. Yet commentators often forget that these alternatives carry their own risks of uncertainty, inherent in the national court review process that can be invoked with respect to any arbitration subject to challenge and enforcement under the New York Convention. Last week’s U.S. court decision in Argentina v. BG Group (D.C. Court of Appeals, No. 1:08-cv-00485) reminds us that whatever arbitral mechanism the parties select, some risk of uncertainty is unavoidable. The debate between ICSID and alternative forums thus should not be framed as one about avoiding uncertainty and promoting finality, but rather about a more fundamental question: who decides?
Much to the surprise of many seasoned international arbitration practitioners, the D.C. Circuit vacated a US$ 185.3 million Final Award against Argentina, essentially nullifying a hard-fought, four-and-a-half year arbitration between the parties. The court vacated the Award on the basis that the “arbitral panel rendered a decision . . . without regard to the contracting parties’ agreement establishing a precondition to arbitration,” namely the clause in the Argentina-UK bilateral investment treaty (BIT) requiring claimants to submit disputes to the Argentine courts for 18 months before resorting to arbitration. In the underlying UNCITRAL arbitration, the tribunal had considered whether the dispute was admissible without having been first submitted to the Argentine courts. It ruled that such submission was not essential because it in this case it would have been an exercise in futility: the claimant could not have obtained relief anyway from the Argentine courts, given the Republic’s apparent interference with access to the courts and its punishment of all would-be local court litigants by excluding them from contract renegotiations. The tribunal concluded that in these circumstances, the 18-month provision could not “be construed as an absolute impediment to arbitration,” and therefore deemed BG Group’s arbitration claims admissible.
By contrast, the D.C. Circuit concluded that this entire analysis was misplaced, since in its view the BIT terms—which it analyzed principally by reference to U.S. domestic law on contractual intent to arbitrate, rather than under the Vienna Convention—were clearly designed to require prior recourse to the Argentine courts. The court found that the tribunal had exceeded its powers by permitting direct access to arbitration contrary to that expressed intent. Indeed, the court suggested that under U.S. case law, the tribunal should not have even engaged in an analysis of the feasibility or usefulness of prior resort to the Argentine courts, because as a threshold matter it had no proper authority under the BIT to admit such issues for substantive consideration.
In the most narrow sense, the D.C. Circuit’s decision did not directly repudiate the years of fairly consistent rulings by ICSID and UNCITRAL tribunals with respect to the 18-month local court requirement under similar Argentine BITs. That is because the BG Group tribunal had not relied on the BIT’s most-favored-nation (MFN) clause, upon which prior tribunals had rested their decisions, even though BG Group did argue that point. Nonetheless, the D.C. Circuit’s analysis implicitly suggests that it also might have overturned an MFN-based decision, since by the Court’s logic, the tribunals who rendered those decisions likewise would have had no authority to bypass the BIT parties’ allegedly clear intent to require local court proceedings in all circumstances. If the decision is read in this broader way, it can be seen as impugning the core logic of many prior decisions. This would include Maffezini v. Spain (ICSID Case No. ARB/97/7, 1 September 2000), where the tribunal allowed an Argentine investor to invoke (by way of an MFN clause) the Chile-Spain BIT to avoid the domestic court prerequisite in the Argentina-Spain BIT; Siemens v. Argentina (ICSID Case No. ARB/028, Decision on Jurisdiction, 3 August 2004), where the tribunal permitted a German investor to invoke the Argentina-Chile BIT to proceed directly to arbitration; National Grid plc v. Argentina (UNCITRAL, Decision on Jurisdiction, 20 June 2006), where the tribunal permitted a British investor to invoke a more favorable term in the Argentina-US BIT to avoid 18 months of litigation in the Argentine courts; and several other cases in the same line. Until the D.C. Circuit’s opinion, the jurisprudence appeared to be converging on consensus regarding the 18-month waiting requirement, even though much controversy remained about the broader application of MFN clauses in other, less procedural, contexts.
Now, with one 17-page decision, a national court not only has completely up-ended the result in one major case, but also in the process unsettled what most observers had thought to be a progression towards certainty, predictability and finality with respect to this issue. Much can—and undoubtedly will— be written about the substance of the court’s analysis. But at heart, it serves as a reminder that some degree of uncertainty is inherent in international arbitration in any forum, so long as there is any mechanism for review and challenge of arbitral awards. This is just as true for the “alternative” routes of ad hoc UNCITRAL or non-ICSID institutional arbitration as it is for ICSID arbitration, since all non-ICSID mechanisms allow for national court challenges under the New York Convention, and national courts (once vested of the matter) may be tempted to apply their own national laws, including on core issues such as arbitrability. Arguably, the uncertainty of national court review may be even greater than that of ICSID annulment review, since most national court judges are comparatively unfamiliar with investment treaty jurisprudence and may be less concerned about contributing to the growth of consensus or emerging doctrine. The choice between the two systems, thus, should not be framed as a quest for predictability and finality, but rather as something more fundamental: a decision about which decision-makers will evaluate challenges, and what rules and standard of review they will use in deciding.
By Jean E. Kalicki and Dawn Yamane Hewett