Rationalizing costs in investment treaty arbitration.

• Author: Franck, Susan D.

TABLE OF CONTENTS I. A PRIMER ON IIAS, ITA, AND RELATED COSTS A. International Investment Agreements (IIAs) B. Investment Treaty Arbitration (ITA) C. Costs of ITA 1. Defining Costs 2. Why Costs Matter 3. Existing Data on ITA II. DOCTRINAL AND POLICY BASES FOR SHIFTING COSTS A. Normative Baselines of Cost Shifting B. Shared Policy Considerations for Cost Shifting C. The Law of Cost Shifting 1. Party Agreement 2. Institutional Rules 3. National Laws 4. International Case Law 5. Sources of Soft Law and Practice III. EMPIRICAL ANALYSIS A. Methodology B. Scope of Cost Decisions C. Hypothesis 1: Descriptive Scope of Cost Decisions 1. Substantive Outcomes of PLC and TCE 2. TCE and PLC in Dollar Values 3. Percentage of PLC and TCE Allocation D. Hypothesis 2: Legal Justification for Costs Decisions 1. The Pre-2007 Data Set: Overall and as a Function of Finality 2. PLC Justifications: Legal Authority and Rationale 3. TCE Justifications: Legal Authority and Rationale E. Hypothesis 3: Relationships Among Costs and Other Variables 1. Relationship Between PLC and TCE 2. Relationship with TCE, Amount Claimed, and Damage Awards 3. The Certainty of Uncertainty IV. ANALYSIS AND RECOMMENDATIONS A. Costs Matter, Need Early Consideration, and Require Additional Data B. Signs of Balance Even Without a "Universal" Approach to Costs C. Gaps in Legal Authority and Rationale Suggest Need for Rationalization D. Links with Cost Variables Suggest Need for Caution and Rationalization CONCLUSION The number of investment treaty arbitrations has nearly quintupled. (1) Billions of dollars--by virtue of cases like the 2002 Argentine currency crisis (2) or the Yukos Oil debacle (3)--are at stake. With global supply chains, massive investment flows, (4) and a network of 2600 treaties, (5) governments are at risk for treaty arbitration when their regulatory measures, like legislation to redress global economic crises, adversely impact foreign investment. (6) Investment treaty arbitration has largely, but not exclusively, been a welcome advance. For foreign investors affected by government conduct, treaty arbitration offers a direct opportunity to sue states and receive damages, whereas alternative venues such as national courts are unavailable or undesirable. (7) Meanwhile, states have an opportunity to protect their investors abroad, vindicate their policy choices, and receive the benefit of increased investment arguably flowing from their investment treaties. (8) Nevertheless, there is a latent problem with investment treaty arbitration, namely, ambiguity about arbitration costs. The scope for cost liability includes: (1) the expenses of both parties' lawyers, (2) the costs of the tribunal and expenses related to administration, and (3) which party will bear these two expenses given the possibility of cost shifting. (9) The scope of cost liability may contribute to concerns about the international investment regime. (10) The lack of certainty and predictability about total costs, which party will have liability for which costs, and the justification for those cost decisions diminishes the effectiveness of investment treaty arbitration. (11)

In Eureko v. Poland, (12) for example, a Dutch investor sued the Republic of Poland under a bilateral investment treaty for problems with the US$1.34 billion insurance privatization. (13) The arbitration made headlines in the international financial news (14) and featured an internationally prominent tribunal. (15) The eighty-six-page award held Poland liable and required Poland to pay the fees of the tribunal and Eureko's lawyers. The arbitrators' full decision on costs was contained in two sentences: "Claimant has prevailed. Consequently, its costs and those of the Tribunal shall be borne by the Respondent." (16) The controlling treaty language prohibited this approach. (17) While the legal error makes it an arguable outlier and a subsequent decision redressed this error, (18) data nevertheless suggests that Eureko's failure to cite any legal authority and the reliance on a single rationale was typical. (19)

Investment arbitration costs are called "a hot issue" (20) and "the sting in the tail." (21) Concerns about the legitimacy of investment treaty arbitration, and incoherency in areas such as costs, may cause states to reevaluate the value of investment treaties. The United States (22) and Norway (23) are reconsidering their model treaties. Meanwhile, Russia withdrew from the Energy Charter Treaty, (24) and Ecuador (25) and Bolivia (26) withdrew from the World Bank's International Centre for the Settlement of Investment Disputes (ICSID). (27)

Better information about investment arbitration costs is necessary. Claims that costs are "no small matter" (28) or may range from US$1-21 million (29) require analysis. Objections that arbitrators "give cursory attention to fixing arbitration costs" (30) require assessment of what justifications tribunals do offer for cost decisions, particularly in the context of investment arbitration. Critiques that cost decisions are "arbitrary" (31) or "unpredictable" (32) necessitate analysis of what variables (if any) are reliably linked to cost decisions. While research in this Article is neither a predictive nor causal model of future outcomes, cost information has the power to (1) aid parties in understanding their arbitration risks and managing their investment treaty disputes, (33) including using Alternative Dispute Resolution (ADR) (34) to facilitate settlement; (35) (2) guide tribunals seeking descriptive data about costs; (36) (3) permit states to design better investment treaties in light of their normative policy choices; (37) and (4) inform the debate about the legitimacy of investment treaty arbitration. (38)

Despite the need for reliable information on investment treaty costs, empirical analysis is just beginning. (39) This Article is the first empirical analysis of investment arbitration costs that recommends potential reforms based upon available data and appropriate norms. Part I of the Article provides a background on investment agreements, treaty arbitration, and costs. Part II explores the doctrinal and normative bases for cost shifting and its application to investment treaty arbitration. Part III describes the methodology, hypotheses, and results of the research.

The results suggested that cost was a key risk in investment treaty arbitration. Even limited data suggested that reported costs represented more than 10% of an average award (i.e., over US$1.2 million). As the data only measured the portion of one party's legal fees that was shifted (and tribunal costs where it was available), it necessarily omitted the full scope of both parties' legal costs; net costs could have been much larger and therefore a more substantial aspect of the amount awarded. (40) Regarding allocation of the risk of liability for arbitration costs, the data showed that costs exhibited a degree of incoherence buttressed by small pockets of coherence. There was no universal approach for how tribunals addressed costs; although tribunals most frequently required parties to share tribunal and administrative costs equally and absorb their own legal fees, there were a mix of approaches and outcomes.

Yet within the variance, the overall experiences of investors and states were relatively equivalent, with (1) parties often responsible for equal costs, or (2) rough parity between investors and states when tribunals did shift costs. There was, however, a lack of justification for these results. Although guidance or decisions on costs could be made at earlier phases, such as in preliminary questions, (41) tribunals typically waited until the end to make decisions. This meant that information, which was possibly vital to strategic settlement opportunities, was unavailable to the parties. When tribunals did make decisions, they did not regularly cite to any legal authority (i.e., citing less than one authority on average) and used minimal justifications (i.e., one to two on average) to justify the result. Where tribunals offered reasons, justifications diverged across categories. Although the literature suggests cost decisions are often based upon a pure "loser-pays" approach or a desire to punish inappropriate behavior, these were not the most frequent rationales; and there was no reliable statistical relationship between losing and cost shifting, either for parties' own legal fees or the tribunal and related administrative costs. (42)

There were other key commonalities. Tribunals were most likely to rationalize (43) their decisions using the parties' relative success and equitable considerations. They were unlikely to base their decisions expressly on concerns related to the public interest, party equality, stare decisis, or settlement efforts. A few remaining areas exhibited a degree of coherence. There was a link between an award's cost decisions, whereby if tribunals shifted attorney's fees onto another party, the same party was also liable for more than 50% of tribunal fees. Finally, there was a reliable relationship between the amounts investors claimed and the tribunal's total costs. If investors made low damage claims, tribunal costs were low; if damage claims were high, tribunal costs were high. As international arbitration has no equivalent to Federal Rule of Civil Procedure 11 requiring good-faith pleadings, the relationship has implications for using cost shifting in arbitration--perhaps even in domestic litigation (44)--to create incentives that promote efficient and fair dispute resolution. Overall, while there were pockets of rough coherence and parity, the larger picture suggests that costs exhibited a degree of uncertainty. The question is whether that is a desirable normative output from a system of justice for international economic law.

In light of these findings and the limitations inherent in the data, measures, and empirical models, Part IV argues...