i Introduction

Bifurcation refers to splitting an arbitration into two separate phases, and most often involves splitting jurisdictional issues from the merits. A tribunal's decision on whether to bifurcate is one of the most critical procedural inflection points in an investment treaty arbitration. If a tribunal bifurcates and dismisses the claims on jurisdiction, the parties avoid briefing the merits, which can save many months (if not years) of time and millions of dollars in legal fees. But, if the tribunal bifurcates and finds jurisdiction, the parties must then embark on a new, separate merits procedure, extending the overall arbitral calendar and increasing costs substantially. It follows that a tribunal's bifurcation decision is a pivotal moment in an investment treaty arbitration.

This chapter focuses on the legal framework that tribunals apply when considering whether to bifurcate jurisdictional objections from the merits (Section II), and the strategic issues parties consider when deciding to seek or to resist jurisdictional bifurcation (Section III). Lastly, we briefly explore the less common practice of bifurcating proceedings to separate the analysis of damages from the merits phase (Section IV).

ii Legal Framework for Considering Applications to Bifurcate Jurisdiction

i The rules governing bifurcation

The two primary sets of procedural rules for investment treaty arbitration – the International Centre for Settlement of Investment Disputes (ICSID) Arbitration Rules and the United Nations Commission on International Trade Law (UNCITRAL) Rules – expressly provide for bifurcation of jurisdictional objections. Article 41(2) of the ICSID Convention states that:

Any objection by a party to the dispute that that dispute is not within the jurisdiction of the Centre, or for other reasons is not within the competence of the Tribunal, shall be considered by the Tribunal which shall determine whether to deal with it as a preliminary question or to join it to the merits of the dispute.

ICSID Arbitration Rule 41 uses similar language, stating in the pertinent part that:

(2) The Tribunal may on its own initiative consider, at any stage of the proceeding, whether the dispute or any ancillary claim before it is within the jurisdiction of the Centre and within its own competence.

. . .

(4) The Tribunal . . . may deal with the objection [that a dispute or any ancillary claim is not within the jurisdiction of the Centre or, for other reasons, is not within the competence of the Tribunal] as a preliminary question or join it to the merits of the dispute.2

The UNCITRAL Rules also permit bifurcation; however, the approach to bifurcation shifted between the 1976 and 2010 versions of the UNCITRAL Rules. Article 24(4) of the 1976 UNCITRAL Rules states that:

In general, the arbitral tribunal should rule on a plea concerning its jurisdiction as a preliminary question. However, the arbitral tribunal may proceed with the arbitration and rule on such a plea in their final award.

Tribunals have held that this language – namely that a tribunal 'should rule' on its jurisdiction 'as a preliminary question' – suggests that the 1976 UNCITRAL Rules create a presumption in favour of bifurcation.3

The approach to bifurcation changed in the 2010 revision to the UNCITRAL Rules. Article 23(3)4 states that:

The arbitral tribunal may rule on a [jurisdictional objection] either as a preliminary question or in an award on the merits.

Tribunals have held that this new formulation – namely that a tribunal 'may rule' on its jurisdiction as a preliminary question – eliminated the presumption in favour of bifurcation that existed in the 1976 UNCITRAL Rules.5 This change brought the UNCITRAL Rules in line with ICSID Arbitration Rules, which likewise include no stated presumption in favour of jurisdictional bifurcation.

ii The legal test for bifurcation

Neither the ICSID Arbitration Rules nor the UNCITRAL Rules provides guidance on the factors that a tribunal should consider when deciding whether to bifurcate proceedings.6 However, investment treaty tribunals have developed and followed a fairly consistent approach to the analysis. Procedural efficiency – whether bifurcation is more likely to increase or decrease the time and costs associated with the arbitration – is the overarching factor that tribunals consider when deciding bifurcation applications. In his seminal treatise on the ICSID Convention, Professor Schreuer explains that the question of whether to bifurcate proceedings is:

a matter of procedural economy. It does not make sense to go through lengthy and costly proceedings dealing with the merits of the case unless the tribunal's jurisdiction has been determined authoritatively. On the other hand, some jurisdictional questions are so intimately linked to the merits of the case that it is impossible to dispose of them in preliminary form.7

Tribunals generally agree with Professor Schreuer's analysis.8 For instance, the Clayton v. Canada tribunal held that 'its purpose in directing bifurcation . . . was to facilitate the efficient litigation of the claim within the arbitration process'.9 In the recent Eco Oro v. Colombia decision on bifurcation, the tribunal held similarly, noting that, in addressing a bifurcation application, 'it should seek to determine what will best serve the Parties and the sound administration of justice, in particular with respect to procedural efficiency'.10 The ICSID Secretariat appears to have embraced this emphasis on efficiency. A new provision appears in the ICSID Secretariat's recently proposed revisions to the ICSID Arbitration Rules, specifying that:

[i]n determining whether to bifurcate, the Tribunal shall consider all relevant circumstances, including whether bifurcation would materially reduce the time and cost of the proceeding.11

To determine whether bifurcation would be efficient, investment treaty tribunals focus their analyses on three questions:12

  1. Is the jurisdictional objection serious and substantial (i.e., does the objection have a reasonable chance of success)?
  2. Will the jurisdictional objection, if sustained, result in a material reduction (or complete elimination) of the merits phase of the proceedings?
  3. Is the jurisdictional objection so intertwined with the merits that bifurcation is impractical?

Each prong of the test is taken up in the sections that follow.

Prong 1: Is the jurisdictional objection serious and substantial?

The first question that tribunals ask when addressing an application to bifurcate is whether the jurisdictional objection is 'serious and substantial'.13 If it is not, the objection is unlikely to succeed, and will not dispense with (or reduce the scope of) the arbitration. It follows that absent one or more serious and substantial objections, a separate jurisdictional phase is needless and wasteful. When addressing whether a jurisdictional objection is serious and substantial at the bifurcation request stage, tribunals do not undertake a full analysis of the objections. Instead, tribunals typically limit their assessment to whether, on its face, the objection is frivolous or clearly without merit. The Resolute Forest tribunal explained the limited nature of the Prong 1 inquiry, noting that:

[t]he determination of the first part of the test, namely whether an objection is prima facie serious and substantial[,] should not, in the Tribunal's view, entail a preview of the jurisdictional arguments themselves. Rather, at this stage the Tribunal is only required to be satisfied that the objections are not frivolous or vexatious.14

This approach creates practical challenges for counsel when crafting Prong 1-related arguments. On the one hand, a respondent15 must set out its arguments in some detail to satisfy the tribunal that its objection is not frivolous. And of course, a claimant will need to assert its counterarguments with some granularity to persuade the tribunal that the respondent's objection is not, in fact, serious and substantial. On the other hand, it is not appropriate for the parties to submit a lengthy, fully formed jurisdictional briefing (backed by extensive jurisprudence, copious factual evidence, etc.) at this stage. For counsel, finding the right balance with respect to Prong 1 can be a challenge.

Prong 2: Will a successful objection materially reduce the scope of the proceedings?

The second question that investment treaty tribunals generally ask when considering a bifurcation application relates to the impact that the jurisdictional objection, if successful, will have on the proceedings. If the objection will not 'materially reduce' the scope of the arbitration, bifurcation will be a wasteful exercise. This of course raises the question: what does 'materially' mean in this context? On one extreme, if a successful jurisdictional objection would end the arbitration, that objection would certainly satisfy Prong 2. This was the situation in the recent Glencore v. Bolivia decision on bifurcation, where the tribunal noted that:

Respondent alleges that the Tribunal lacks jurisdiction because Glencore Bermuda committed an abuse of process by structuring an investment in order to obtain standing. . . . as to [this objection] it is clear that, if successful, these proceeding [sic] would be brought to an end.16

However, if a successful objection would only narrow, but not eliminate, a future merits phase, this may not be sufficient to warrant bifurcation. The Gavrilovic v. Croatia tribunal held that a bifurcation application failed Prong 2 of the analysis because:

even if the Respondent were successful in its [jurisdictional argument], it would not appear to obviate the need for a merits phase. . . . Indeed, the scope of the dispute, although perhaps narrower, may not be so narrowed as to warrant the cost, expense and inconvenience of dividing the proceeding into two phases. Put simply, as the Claimants contend, it appears not to be a substantial enough objection, in and of itself, to justify bifurcation.17

As this analysis illustrates, the requirement that a jurisdictional objection can materially reduce the scope of the arbitration is a sensible one. It would be inefficient to hear a jurisdictional objection in a separate phase if, regardless of the outcome, the parties will nonetheless proceed to a complex and costly merits phase of the proceedings.

Prong 3: Is the jurisdictional objection intertwined with the merits?

The third question that investment treaty tribunals typically ask when considering bifurcation is whether reaching a decision on the jurisdictional objection will require an examination of the merits of the case. If the jurisdictional arguments are too intertwined with merits-related issues, tribunals generally refuse to bifurcate.18 Professor Schreuer confirms that bifurcation is not appropriate 'where the answer to the jurisdictional questions depends on testimony and other evidence that can only be obtained through a full hearing of the case'.19

The Gavrilovic v. Croatia tribunal explored Prong 3 in detail, ultimately concluding that the jurisdictional objections in that case were not sufficiently separated from the merits.20 The Gavrilovic tribunal held that:

a ruling on at least three of the four preliminary objections would in all likelihood require a detailed examination of the same evidence that will ultimately need to be examined at the stage of determining the merits. There is no procedural or other advantage with bifurcating the proceeding, so as to require not only the Tribunal to consider the same, or similar, evidence on two occasions, but so as to require witnesses to appear on two occasions, submissions to be prepared which canvass the same, or similar, matters, and the consequential cost and expense. . . . Once a considerable factual overlap is accepted, which the Tribunal considers to be the case, little can be said in support of the division of the case.21

As the Gavrilovic tribunal explained clearly, there is little sense in bifurcating a jurisdictional objection if, to decide that objection, a tribunal would need to examine anyway the merits of the claimant's claims. Furthermore, tribunals have held that addressing merits-related issues at the jurisdictional stage – before the tribunal can review all of the evidence and arguments on the merits – could raise due process issues related to 'prejudging' the merits.22

The three-part analysis described above is a conjunctive test. Tribunals typically only favour bifurcation when a preliminary objection meets all three criteria. It would seem ill-advised, for example, to bifurcate an objection that was not serious and substantial (Prong 1), even if the objection would dispose of the case if successful and was distinct from the merits (Prongs 2 and 3). Likewise, it would appear unwise to bifurcate an objection that would not dispose of 'all or an essential part of' the claims (Prong 2), even if the objection were serious and substantial and distinct from the merits (Prongs 1 and 3). For these reasons, tribunals typically deny bifurcation if an objection fails even one of the three prongs of the test.

III Strategic Considerations Regarding Bifurcation Applications

In light of the significant impact that bifurcation can have on the arbitral process, parties should consider carefully whether to request or to resist bifurcation. Unsurprisingly, the strategic considerations for claimant investors and respondent states differ, as are described below.

i Strategic considerations for respondents

The conventional wisdom is that if a respondent state has reasonably sound jurisdictional objections, the respondent should seek to bifurcate the proceedings. There is logic to this approach: if a respondent prevails on one or more of its jurisdictional objections, this could dispense with, or at least lessen the complexity of, the merits phase of the arbitration. In addition to this clear upside, there are at least three other potential benefits to a respondent seeking bifurcation.

First, bifurcated proceedings often focus the tribunal's attention (at least initially) on facts favourable to the respondent. In many cases, the factual story related to a jurisdictional objection (be it corruption, illegality, abuse of process, etc.) does not present a claimant in a positive light. However, the facts related to the merits (e.g., alleged expropriation or unfair and inequitable government treatment) may well paint a less-than-ideal picture of the respondent state. In that scenario, by isolating jurisdictional issues from the rest of the case, a respondent can focus the tribunal's attention on factual issues that are unfavourable to the claimant, while keeping its own alleged misdeeds largely out of the spotlight. By shaping and narrowing the story in this way, the respondent gains a strategic advantage.

Second, respondents might seek to use bifurcation to extend arbitral proceedings to delay any possible adverse consequences of their actions. Although this is an approach the authors do not endorse, commentators have observed that '[a] party may be advocating bifurcation to delay and obstruct the arbitration, rather than to make it more efficient.'23 The appeal of this strategy is not difficult to discern: in a bifurcated proceeding, even if the jurisdictional claim fails, the respondent succeeds in extending the arbitral calendar, and thus, delaying its receipt of a potential adverse award. (Though a tribunal's eventual award of interest could in theory nullify any financial advantage of such a delay, financial motives may not be the only ones in play.)

Third, if a respondent state presents a jurisdictional objection, but does not seek bifurcation, a tribunal might (rightly or wrongly) perceive this as an indication that the respondent believes its jurisdictional objection is weak. It could be argued that if a respondent truly believes its jurisdictional objection will succeed, it will seek to have that objection heard before arguing the merits.

There are, however, perfectly valid reasons why a respondent state might not want to seek bifurcation. For example, if a respondent believes that the overall equities favour the state – imagine, for example, that the respondent has evidence that the investor caused serious environmental damage that it wants the tribunal to see – it might behoove that respondent to tell its whole story (jurisdiction and merits) all at once for maximum impact. Or, if a respondent's jurisdictional objection will not significantly streamline the case if successful, the respondent might be best served by not seeking bifurcation, thereby avoiding additional costs. More often, however, if a respondent has strong jurisdictional objections that will have a material impact on the scope of the case, seeking bifurcation is a prudent course.

ii Strategic considerations for claimants

Claimants typically oppose bifurcation, and do so for reasons that are often the converse of the factors that lead respondents to lodge bifurcation requests. First, claimants generally prefer to avoid stand-alone jurisdictional phases if they will focus on the claimants' own misconduct (e.g., corruption or abuse of rights), while deferring consideration of the compelling facts that claimants want to highlight (e.g., facts related to expropriation or unfair treatment). In those cases, claimants will perceive a benefit in presenting their merits-related story at the same time that they argue jurisdiction.

Second, unlike a respondent (which might prefer for various reasons to extend the proceedings), a claimant's incentive is to conclude the arbitration and receive a favourable award as quickly as possible. The most direct path to that outcome is an arbitral process in which the tribunal hears jurisdiction and merits together.

Third, Prong 1 of the bifurcation analysis relates to whether the jurisdictional objection is serious and substantial. Thus, an application for bifurcation necessarily involves an argument from the respondent that it is asserting a serious and substantial objection. If a claimant does not challenge that characterisation, a tribunal might interpret this as a sign that the respondent's objections have traction. At the very least, by not opposing bifurcation, a claimant forgoes an early opportunity to explain to the tribunal why the respondent's jurisdictional objection is not, in fact, serious and substantial.

For these reasons, it is typically advisable for claimants to oppose bifurcation. As always, however, there are exceptions to the rule. For instance, if a respondent raises a jurisdictional objection that is indeed distinct from the merits and has the potential to dispose of the case, it might make sense for a claimant to agree to bifurcate that issue to save its own resources as well. The case for doing so may be even stronger if, in exchange for agreeing not to oppose bifurcation, the claimant can secure the respondent's agreement to an accelerated procedural calendar for the jurisdictional phase. Imagine, for example, a hypothetical case where a respondent state objects to ICSID jurisdiction based on its withdrawal from the ICSID Convention. This is a purely legal issue that is wholly separate from the substance of the hypothetical investor's claim, and may be amenable to a relatively speedy resolution. In this scenario, it might be wise for a claimant to agree to obtain the tribunal's verdict on the threshold issue before expending the resources to fully develop and argue its case on the merits.

iv Bifurcation of Damages

Historically, the vast majority of bifurcated proceedings have involved the separation of the jurisdictional and merits phases of an arbitration. This section, however, briefly discusses a different category of bifurcation: separating the damages phase of the proceedings from the merits.24 The rules of the major arbitral institutions do not specifically provide for the bifurcation of damages or quantum issues. However, tribunals typically cite procedural rules granting them general authority to control the arbitration procedure as a basis for bifurcating damages.25

Although bifurcating damages is not common, it also not unheard of – tribunals do, on occasion, decide damages-related issues separately from the merits, and considerations of efficiency appear to predominate in their analysis of the decision to do so, just as in the bifurcation of jurisdiction and merits.26 For instance, in Suez v. Argentina, the tribunal decided that the most efficient approach would be to separate its damages analysis from its decision on the liability of the respondent. The Suez tribunal stated that it:

has decided to render a decision on liability before arriving at an award on damages. It has chosen to adopt this procedure for reasons of judicial economy. Given the complexity of this case and the extraordinarily voluminous nature of the record, the Tribunal by rendering a decision on liability now and thereby defining the scope of its investigation with respect to a determination of damages will be able more efficiently to define the mission of the independent expert that will assist the Tribunal in this determination.27

The tribunal in Glencore v. Bolivia took a similar approach. In that case, the tribunal rejected an application to hear jurisdictional challenges in a separate phase, but chose to bifurcate damages, noting that '[t]his approach seems to the Tribunal more efficient in terms of time and costs than the alternative'.28

Bifurcating the damages phase is more likely to produce efficiencies in particular situations. For instance, where a claimant plans to assert particularly complex, novel or fact-intensive damages arguments, bifurcation of damages may be warranted for at least two reasons. First, if the claims fail at the jurisdictional or merits stage, both parties would avoid the time and expense associated with developing their complicated damages-related theories. Second, a separate damages phase would ensure that parties have sufficient opportunity to focus on, and fully develop, their damages theories and counterarguments. Another scenario that might warrant bifurcation of damages would be a case involving jurisdictional objections with the potential to eliminate some claims, particularly if the damages are expected to differ from claim to claim or overlap in complex ways across claims. In that situation, where the jurisdictional outcome may significantly reduce the scope or complexity of the damages analysis, it may make sense to bifurcate and address damages only after it is clear which claims are left standing.

v Conclusion

As this chapter highlights, a decision on bifurcation is a critical procedural moment in an arbitration. Though bifurcation can accelerate the resolution of a dispute, it can also significantly prolong the arbitral process. For this reason, tribunals have developed – and have applied with commendable consistency – the detailed, tripartite test described above. The parties also play an important role. Their strategic decisions regarding whether to request or to oppose bifurcation will shape the arbitration in important ways, impacting the length and cost of the arbitral proceedings.


Footnotes

1 Marinn Carlson is a partner and Patrick Childress is a senior associate at Sidley Austin LLP.

2 The ICSID Arbitration Rules also provide that '[u]pon the formal raising of an objection relating to the dispute, the Tribunal may decide to suspend the proceeding on the merits.' ICSID Arbitration Rules, Rule 41(3). However, the ICSID Secretariat recently proposed amendments to the ICSID Arbitration Rules that would remove tribunals' power to suspend the proceedings during the pendency of a bifurcation application. See Proposals for Amendment of the ICSID Rules, Consolidated Draft Rules, Volume II, August 2, 2018, Rule 36.

3 See, e.g., Mesa Power Group LLC v. Government of Canada, PCA Case No. 2012-17, Procedural Order No. 2, January 18, 2013, para. 16 (holding that under the 1976 UNCITRAL Rules 'when a Party raises an objection to jurisdiction, the presumption is in favor of addressing the objection as a preliminary question'); President Allende Foundation, Victor Pey Casado and Coral Pey Grebe v. Republic of Chile II, PCA Case No. 2017-30 (PCA Case No. AA662), Procedural Order No. 2, November 29, 2017, para. 63.

4 This provision is unaffected by the 2013 revisions to the UNCITRAL Rules (incorporating rules on transparency in investment treaty arbitration).

5 See, e.g., Glencore Finance (Bermuda) Limited v. Plurinational State of Bolivia, PCA Case No. 2016-39, Procedural Order No. 2 (Decision on Bifurcation), January 31, 2018, para. 38; Resolute Forest Products Inc. v. Government of Canada, PCA Case No. 2016-13, Procedural Order No. 4 re Decision on Bifurcation, November 18, 2016, para. 4.3.

6 See, e.g., Global Telecom Holding S.A.E. v. Canada, ICSID Case No. ARB/16/16, Procedural Order No. 2 Decision on Respondent Request for Bifurcation, December 14, 2017 (holding that '[n]either the ICSID Convention nor the Arbitration Rules sets forth a legal standard applicable to the decision of whether to join preliminary objections to the merits or instead to hear them in a preliminary phase. The ICSID Convention and the Arbitration Rules leave this decision entirely to the discretion of tribunals').

7 C. Schreuer, The ICSID Convention: A Commentary, 2nd Edition, p. 537.

8 Lighthouse Corporation Pty Ltd and Lighthouse Corporation Ltd, IBC v. Democratic Republic of Timor-Leste, ICSID Case No. ARB/15/2, Procedural Order No. 3 on Bifurcation and Related Requests, July 8, 2016, para. 19; William Ralph Clayton and others v. Government of Canada, PCA Case No. 2009-04, Procedural Order No. 19, August 10, 2015, para. 19.

9 William Ralph Clayton and others v. Government of Canada, PCA Case No. 2009-04, Procedural Order No. 19, August 10, 2015, para. 19.

10 Eco Oro Minerals Corp. v. Republic of Colombia, ICSID Case No. ARB/16/41, Procedural Order No. 2, June 28, 2018, para. 50.

11 Proposals for Amendment of the ICSID Rules, Consolidated Draft Rules, Volume II, August 2, 2018, Rule 37(4).

12 This three-part test was first set out in Glamis Gold v. Canada. See Glamis Gold, Ltd., v. United States of America, UNCITRAL, Procedural Order No. 2 (Revised), May 31, 2005, para. 12. Since Glamis, many other tribunals have adopted the same, or a very similar, approach. See, e.g., Philip Morris Asia Limited v. The Commonwealth of Australia (UNCITRAL) Procedural Order No. 8 Regarding Bifurcation of the Procedure, April 14, 2014, para. 109; Emmis International Holding, B.V., Emmis Radio Operating, B.V. and MEM Magyar Electronic Media Kereskedelmi és Szolgáltató Kft. v. Hungary, ICSID Case No. ARB/12/2, Decision on Application for Bifurcation, June 13, 2013, para. 37; A11Y Ltd. v. Czech Republic, ICSID Case No. UNCT/15/1, Procedural Order No. 2 Decision on Bifurcation, October 5, 2015, paras. 56-58; Michael Ballantine and Lisa Ballantine v. Dominican Republic, PCA Case No. 2016-17, Procedural Order No. 2, April 21, 2017, para. 18; Cairn Energy PLC and Cairn UK Holdings Limited (CUHL) v. Republic of India, PCA Case No. 2016-7, Decision on the Respondent Application for Bifurcation (Procedural Order No. 4), April 19, 2017, paras. 76-77; Glencore Finance (Bermuda) Limited v. Plurinational State of Bolivia, PCA Case No. 2016-39, Procedural Order No. 2 (Decision on Bifurcation), January 31, 2018, para. 39; Eco Oro Minerals Corp. v. Republic of Colombia, ICSID Case No. ARB/16/41, Procedural Order No. 2, June 28, 2018, para. 49.

13 See, e.g., Lighthouse Corporation Pty Ltd and Lighthouse Corporation Ltd, IBC v. Democratic Republic of Timor-Leste, ICSID Case No. ARB/15/2, Procedural Order No. 3 on Bifurcation and Related Requests, July 8, 2016, para. 23 (holding that Timor Leste's objections regarding consent to ICSID arbitration, the existence of a cognizable investment, and the alleged foreign investor's nationality 'do not appear frivolous', and thus, satisfy Prong 1 of the test).

14 Resolute Forest Products Inc. v. Government of Canada, PCA Case No. 2016-13, Procedural Order No. 4 re Decision on Bifurcation, November 18, 2016, para. 4.4.

15 Of course, a jurisdictional objection may also be raised by a claimant against a respondent state's ancillary claim (such as a counterclaim) – and thus, at least in theory, bifurcation could equally be sought by a claimant looking to dispose of a respondent state's claim. For convenience, however, the discussion here will use terminology associated with the far more common scenario of a respondent state proposing bifurcation and the claimant investor resisting it.

16 Glencore Finance (Bermuda) Limited v. Plurinational State of Bolivia, PCA Case No. 2016-39, Procedural Order No. 2 (Decision on Bifurcation), January 31, 2018, paras. 43-44.

17 Gavrilovic and Gavrilovic d.o.o. v. Republic of Croatia, ICSID Case No. ARB/12/39, Decision on Bifurcation, January 21, 2015, para. 82.

18 See, e.g., Gavrilovic and Gavrilovic d.o.o. v. Republic of Croatia, ICSID Case No. ARB/12/39, Decision on Bifurcation, January 21, 2015, para. 93. See also Lighthouse Corporation Pty Ltd and Lighthouse Corporation Ltd, IBC v. Democratic Republic of Timor-Leste, ICSID Case No. ARB/15/2, Procedural Order No. 3 on Bifurcation and Related Requests, July 8, 2016, para. 25.

19 C. Schreuer, The ICSID Convention: A Commentary, 2nd Edition, pp. 538-39; Standard Chartered Bank (Hong Kong) Limited v. United Republic of Tanzania, ICSID Case No. ARB/15/41, Procedural Order No. 3 on Bifurcation, October 11, 2016, para. 56; Abertis Infraestructuras, S.A. v. Argentine Republic, ICSID Case No. ARB/15/48, Procedural Order No. 2, March 27, 2017, para. 40; Michael Ballantine and Lisa Ballantine v. Dominican Republic, PCA Case No. 2016-17, Procedural Order No. 2, April 21, 2017, para. 28.

20 Gavrilovic and Gavrilovic d.o.o. v. Republic of Croatia, ICSID Case No. ARB/12/39, Decision on Bifurcation, January 21, 2015, para. 93.

21 ibid.

22 See Glencore Finance (Bermuda) Limited v. Plurinational State of Bolivia, PCA Case No. 2016-39, Procedural Order No. 2 (Decision on Bifurcation), January 31, 2018, paras. 39, 53; Global Telecom Holding S.A.E. v. Canada, ICSID Case No. ARB/16/16, Procedural Order No. 2 Decision on Respondent Request for Bifurcation, December 14, 2017, para. 100.

23 L. Greenwood, 'Does Bifurcation Really Promote Efficiency?', 28:2 J. Int'l Arb. 105, 107 (2011).

24 We use 'the merits' here to encompass all stages of the case leading to a finding of liability, whether or not those prior stages include jurisdictional objections (or indeed, even bifurcation of jurisdictional issues, which can potentially lead to 'trifurcation' of the case into jurisdiction, merits, and damages phases).

25 For instance, in Suez v. Argentina, the tribunal based its authority to bifurcate damages on Article 44 of the ICSID Convention, which states that: '[i]f any question of procedure arises which is not covered by this Section or the Arbitration Rules or any rules agreed by the parties, the Tribunal shall decide the question'. See Suez, Sociedad General de Aguas de Barcelona S.A. and Vivendi Universal S.A v. Argentine Republic, ICSID Case No. ARB/03/19, Decision on Liability, July 30, 2010, para. 272.

26 See, e.g., Suez, Sociedad General de Aguas de Barcelona S.A. and Vivendi Universal S.A v. Argentine Republic, ICSID Case No. ARB/03/19, Decision on Liability, July 30, 2010, para. 272; Glencore Finance (Bermuda) Limited v. Plurinational State of Bolivia, PCA Case No. 2016-39, Procedural Order No. 2 (Decision on Bifurcation), January 31, 2018, para. 56.

27 See, e.g., Suez, Sociedad General de Aguas de Barcelona S.A. and Vivendi Universal S.A v. Argentine Republic, ICSID Case No. ARB/03/19, Decision on Liability, July 30, 2010, para. 272.

28 Glencore Finance (Bermuda) Limited v. Plurinational State of Bolivia, PCA Case No. 2016-39, Procedural Order No. 2 (Decision on Bifurcation), January 31, 2018, para. 56.