523
SGS v. Philippines and the Role of ICSID
Tribunals in Investor-State Contract Disputes
MATTHEW WENDLANDT
*
SUMMARY
I. INTRODUCTION.......................................................................................................524
II. A BRIEF HISTORY OF PRIVATE INTERNATIONAL INVESTMENT ...............................526
A. Bilateral Investment Treaties (BITs) .............................................................526
B. International Centre for the Settlement of Investment Disputes (ICSID)......529
1. A Forum for the Resolution of BIT Claims.............................................529
2. The Source of ICSID Jurisdiction...........................................................530
3. The Effect of Umbrella Clauses ..............................................................532
III. BIT CLAIMS AND CONTRACT FORUM-SELECTION CLAUSES ..................................533
A. Lanco v. Argentina ........................................................................................533
B. Vivendi v. Argentina .....................................................................................534
IV. THE SGS CASES: OF CONTRACT CLAIMS AND FORUM-SELECTION CLAUSES ........537
A. Declining Jurisdiction over Contract Claims: The Approach in SGS v.
Pakistan .........................................................................................................537
1. Background.............................................................................................537
2. Analysis...................................................................................................538
B. Accepting Jurisdiction Over Contract Claims: The Approach of SGS v.
Philippines.....................................................................................................542
1. Background.............................................................................................542
2. Analysis...................................................................................................546
V. EVALUATING THE ROLE OF ICSID TRIBUNALS: THE SHORTCOMINGS OF A
―DECIDE EVERYTHING‖ APPROACH .......................................................................549
A. Consent Controls: The BIT as an Agreement to Arbitrate ...........................549
B. Certainty in an Uncertain World? Choices Facing ICSID Tribunals ..........551
VI. CONCLUSION..........................................................................................................555
* The University of Texas School of Law, J.D. expected May 2008.
524 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
I. INTRODUCTION
Few areas of international law have undergone more change in the last half-century
than the relationship between investors and the foreign countries in which they invest.
Bilateral investment treaties (BITs), which govern the relationship between investors of one
state and the government of another, have been increasingly prevalent since the 1960s.1
In
addition to providing a number of substantive protections to international investors, many
BITs provide for investor-state arbitration before the International Center for the Settlement
of Investment Disputes (ICSID).2
Established by the World Bank in 1966, ICSID provides
an arbitration and conciliation forum to all states who are signatories to the ICSID
Convention.3
Dispute resolution clauses in BITs overwhelmingly provide for ICSID
arbitration, either exclusively or as an available alternative.4
Consequently, the increase of
BITs has seen a corresponding surge in ICSID arbitrations.5
The stakes in this type of
arbitration are substantial, with most awards constituting hundreds of millions of dollars.6
ICSID is a forum that resolves foreign investment disputes, defined as disputes
―between an investor from one country and a government that is not its own that relat[e] to
an investment in the host country.‖7
It is important to note that today ICSID arbitration
generally arises in one of two ways.
In the first situation, an investor who has contracted with another private party in the
host country claims that the host country has breached substantive provisions of the BIT.
For example, the investor may claim that the host country expropriated an investment
without compensation or did not afford the investor fair and equitable treatment. Thus, in
addition to any private cause of action against the private party, an investor has a claim
1. RUDOLF DOLZER & MARGRETE STEVENS, BILATERAL INVESTMENT TREATIES xii (1995).
2. See U.N. CONF. ON TRADE AND DEV. (―UNCTAD‖), BILATERAL INVESTMENT TREATIES,
1959-1999, at 20, U.N. Doc. UNCTAD/ITE/IIA/2 (2000); UNCTAD, BILATERAL INVESTMENT TREATIES
IN THE MID-1990s, at 94–95, U.N. Doc. UNCTAD/ITE/IIT/7 (1998). See generally UNCTAD, New York,
Geneva, 2003, Course on Dispute Settlement in International Trade, Investment and Intellectual Property:
International Centre for Settlement of Investment Disputes: 2.1 Overview, U.N. Doc. UNCTAD/EDM/Misc.232
(prepared by Christoph Schreuer) [hereinafter Course on Dispute Settlement].
3. Convention on the Settlement of Investment Disputes Between States and Nationals of Other States,
opened for signature Mar. 18, 1965, 17 U.S.T. 1270 (entered into force Oct. 14, 1996) [hereinafter ICSID
Convention]; Course on Dispute Settlement, supra note 2, at 13, 15. ICSID also serves as a potential arbitration
forum for non-signatory states. See ICSID Additional Facility Rules,
http://www.worldbank.org/icsid/facility/AFR_English-final.pdf (last visited Apr. 19, 2008).
4. See LUCY REED ET AL., GUIDE TO ICSID ARBITRATION 4 (2004) (―[M]any, if not most BITs, include the
option of ICSID dispute resolution.‖); Christoph Schreuer, Travelling the BIT Route: Of Waiting Periods,
Umbrella Clauses and Forks in the Road, 5 J. WORLD INVESTMENT & TRADE 231, 231 (2004) (―Most [BITs] refer
to ICSID.‖). See also The Treaty Between the United States of America and the Republic of Cameroon
Concerning the Reciprocal Encouragement and Protection of Investments, art. 7(3) Feb. 26, 1986, S. TREATY
DOC. NO. 99-22, [hereinafter U.S.-Cameroon BIT] (consenting to submit investment disputes exclusively to
ICSID); The Treaty Between the United States of America and the Argentine Republic Concerning the Reciprocal
Encouragement and Protection of Investment, U.S.-Arg., art. VII (4), Nov. 14, 1991, S. TREATY DOC. No. 103-2
(consenting in advance to any one of the available forms of arbitration provided in the BIT, including but not
limited to ICSID arbitration).
5. Judith Gill et al., Contractual Claims and Bilateral Investment Treaties: A Comparative Review of the
SGS Cases, 21 J. INT‘L. ARB., 397, 397 n.4 (2004) (―Of the 63 cases pending before the Centre during the year
[2003], 47 were submitted under the dispute-settlement provisions of bilateral investment treaties‖).
6. See Susan D. Franck, The Legitimacy Crisis in Investment Treaty Arbitration: Privatizing Public
International Law through Inconsistent Decisions, 73 FORDHAM L. REV. 1521, 1521 n.1 (2005) (detailing claims
seeking millions of dollars against sovereign states for breaches of international investment agreements).
7. R. DOAK BISHOP, JAMES CRAWFORD & W. MICHAEL REISMAN, FOREIGN INVESTMENT DISPUTES: CASES,
MATERIALS AND COMMENTARY 9 (2005).
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 525
based on international law.8
In effect, the investor is initiating arbitration proceedings to
which the host country has already consented, either in the BIT or in the host country‘s
national law.9
In the second situation, a dispute arises after the investor signs a contract with the host
country itself or with an agency of the host country. This type of foreign investment dispute
is the article‘s focus. Just as in the case involving a private contract, the investor is able to
bring BIT claims against the host government under the BIT‘s substantive provisions. In
addition, the investor may sue in contract against the host country. Therefore, two types of
claims against the host country potentially exist: treaty-based claims and contract-based
claims. The host country‘s status as a party to both the BIT and the contract poses a
problem for ICSID tribunals determining whether they have jurisdiction.
ICSID tribunals derive their jurisdiction solely from the consent given by countries in
the BIT.10
While the BIT provides for arbitration in a neutral forum such as ICSID, the
investor-state contract usually contains a forum-selection clause specifying the courts or
tribunals of the host country.11
It is now widely accepted that ICSID tribunals have
jurisdiction over BIT claims notwithstanding a forum-selection clause in the contract.12
But
in recent years ICSID tribunals have reached opposite conclusions regarding whether a
tribunal has jurisdiction to decide an investor‘s contract-based claims.
The tribunals in SGS v. Pakistan and SGS v. Philippines were presented with the same
issue: whether an ICSID tribunal has jurisdiction over an investor‘s contract claims despite
an exclusive forum-selection clause in the contract.13
In each case, the investor pointed to a
provision of the BIT referred to as an ―umbrella clause‖ and argued that the contract claim
was transformed by the clause into a BIT claim over which ICSID had jurisdiction. Of the
two tribunals, only the tribunal in SGS v. Philippines decided that it had jurisdiction and
found itself in a position to confront a second issue: after accepting jurisdiction over a
8. Id. at 837 (―Customary international law has long provided for the protection of aliens, including the
protection of their investments in the territory of other states . . . BITs may play some role in the formation of
customary international law. Hence, there may be certain overlaps . . . .‖).
9. Id. at 317 (―ICSID‘s jurisdiction may now be secured by a unilateral statement on behalf of a contracting
state party to the Washington Convention and may, also, be incorporated in BITs as well as NAFTA.‖);
International Centre for Settlement of Investment Disputes (ICSID) Official website,
http://www.worldbank.org/icsid/about/about.htm (last visited Oct. 16, 2007) [hereinafter ICSID Official website]
(―Advance consents by governments to submit investment disputes to ICSID arbitration [are] found in about
twenty investment laws and in over 900 bilateral investment treaties.‖). See Lanco Int‘l, Inc. v. Argentine
Republic, Preliminary Decision on Jurisdiction, ICSID Case No. ARB/97/6, para. 32 (1998) (―The States that are
signatories to the [BIT] make a generic offer . . . to submit to international arbitration as selected by the investor
. . . .‖).
10. ICSID Convention, supra note 3, art. 25(1). Again, consent may also be given unilaterally, that is,
outside of a treaty as part of national law. See ICSID Official website, supra note 9. See also DOLZER &
STEVENS, supra note 1, at 131 (―The consent of the two parties, both the governmental party and the investor, has
been called the ‗cornerstone‘ of ICSID‘s jurisdiction.‖).
11. Of course, the investor would rather avoid the prospect of a partisan forum, but the highly competitive
nature of foreign investment strengthens the host country‘s bargaining position, and the result is almost always a
host-friendly forum-selection clause. See, e.g., SGS Société Générale de Surveillance S.A. v. Islamic Republic of
Pakistan, Decision of the Tribunal on Objections to Jurisdiction, ICSID Case No. ARB/01/13, para. 157 (2002)
(providing for dispute resolution in a forum other than a Pakistani arbitration tribunal would have been a ―dealbreaker‖
for host country Pakistan.).
12. Lanco, ICSID Case No. ARB/97/6 para. 32; Vivendi v. Argentina, Decision on Annulment, ICSID Case
No. ARB/97/3, para. 14 (2002).
13. SGS v. Pakistan, ICSID Case No. ARB/01/13 para. 157; SGS Société Générale de Surveillance S.A. v.
Republic of the Philippines, ICSID Case No. ARB/02/6, paras. 16–17 (2004).
526 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
contract claim, must an ICSID tribunal decide the merits of the dispute?14
The tribunal held
that it had jurisdiction, but it did not decide the merits, choosing instead to give effect to the
contract‘s forum-selection clause and stay the proceedings.
Current commentary criticizes the inconsistency of the SGS cases, sometimes using the
disparate results as a platform to advocate greater consistency among ICSID tribunals or
among international arbitration tribunals in general.15
Others go a step further, saying that
Philippines was right to accept jurisdiction but wrong for stopping short of deciding the
merits of the underlying claim.16
Needed at present is a comparison of the SGS decisions
that examines whether either outcome is consistent with other ICSID decisions and with
international legal policies. In addition, SGS v. Philippines seems to upset the prevailing
assumption that if jurisdiction is accepted, an ICSID tribunal must decide the merits of the
case in order to safeguard investors‘ expectations and provide certainty.
This article argues that of the two SGS cases, Philippines finds the better approach by
accepting jurisdiction over contract claims. Furthermore, the tribunal‘s decision to give
effect to the contract‘s forum-selection clause is in accord with both previous ICSID
decisions and the policies underlying private international investment law. Although there
are circumstances in which a tribunal might find it appropriate to decide the contract-based
claim, safeguarding investment expectations is an inadequate justification for deciding
contract claims in each and every case.
Part One of the paper provides a brief history of private international investment law,
focusing on bilateral investment treaties and ICSID arbitration. Part Two lays essential
groundwork for the SGS cases by discussing two decisions, Lanco v. Argentina and Vivendi
v. Argentina. Both cases define the scope of ICSID‘s jurisdiction over BIT claims, and
Vivendi provides a test for distinguishing treaty-based and contract-based claims that is
employed by both SGS tribunals. Part Three explores the SGS cases in two segments: first,
the tribunals‘ different approaches to the question of jurisdiction over contract claims, and
second, the SGS v. Philippines decision to accept jurisdiction but leave resolution of the
merits to the forum designated by the contract. Part Four examines the result in Philippines
to determine its congruity with international investment policies. Part Five concludes that
the Philippines result gives ICSID tribunals the latitude to which they are entitled.
II. A BRIEF HISTORY OF PRIVATE INTERNATIONAL INVESTMENT
A. Bilateral Investment Treaties (BITs)
The lure of undeveloped and developing markets in foreign countries has long held
attraction for intrepid investors. Along with the potential for reward, however, came
significant risk. Expropriation by a host state without adequate compensation could mean a
14. See Ignacio Suarez Anzorena, et al., International Legal Development in Review: 2005 Disputes, 40
INT‘L LAW. 251, 255 (2005) (framing the issues as follows: ―[C]an an umbrella clause transform a contract breach
into a treaty breach; and if it can, must the breach be of a particular character or magnitude to engage the host
state‘s responsibility?‖).
15. See Franck, supra note 6, at 1568-74; David A. Gantz, An Appellate Mechanism for Review of Arbitral
Decisions in Investor-State Disputes: Prospects and Challenges, 39 VAND. J. TRANSNAT‘L L. 39, 49 (2006);
Yuval Shany, Contract Claims vs. Treaty Claims: Mapping Conflicts Between ICSID Decisions on Multisourced
Investment Claims, 99 AMER. J. INT‘L L. 835 (2005).
16. See, e.g., Jarrod Wong, Umbrella Clauses in Bilateral Investment Treaties: Of Breaches of Contract,
Treaty Violations, and the Divide Between Developing and Developed Countries in Foreign Investment Disputes,
14 GEO. MASON L. REV. 135 (2006).
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 527
complete loss for the investor despite carefully laid plans. In the past, the investor would
petition his own country to flex its military might, bullying the less developed country into
capitulating and refunding the investor‘s money.17
Such unwieldy ―gunboat diplomacy‖
made few investors‘ prospects secure, for they had no assurance that their government
would find their particular grievance a worthy cause to support and there was no guarantee
of full recovery. For the developing capital-importing country, resentment against foreign
investment and the conflicts it engendered became widespread. The Calvo Doctrine, which
claims that foreign investors are not entitled to equal or better treatment than the host
country‘s own citizens, requires foreign investors to seek redress in the country‘s courts and
nowhere else.18
Although such hostility towards foreign investors still finds a voice today,19
in general, host countries‘ need for capital and investors‘ desire for profit have compelled
agreement and compromise.
The mutually beneficial effects of foreign investment became increasingly apparent
following the Second World War, as market forces encouraged increased flow of capital
among investors and developing countries.20
As a result of the socialization of Eastern
Europe, increased nationalization of certain industries worldwide, and newly freed colonial
territories, opportunities for investment were matched by an increasing probability of
investment disputes.21
By this time, although gunboat diplomacy had become a relic of the
past, investors still had little protection save for ―the cumbersome and politically costly
procedure of diplomatic protection.‖22
Even investors whose countries had treaties in place to govern economic relations with
other sovereigns found the existing structure inadequate for the post-WWII frontier. For
example, the United States had come to rely on Friendship, Commerce, and Navigation
treaties (FCNs) to govern a wide range of economic issues with foreign countries.23
During
the 1950s, other nations adopted FCNs also.24
But, since they were generally signed
between two developed governments, FCNs were ill-equipped to balance competing
interests between developed and developing countries. Specifically, developing countries
were unwilling to grant foreign investors the rights to which FCNs entitled them, such as
unrestricted entry and unqualified equal treatment.25
When conflicts arose, FCNs provided
no ready mechanism for resolution.26
It became increasingly apparent that a new type of
treaty to govern investor-state relations was needed.
17. See, e.g., HOLGER H. HERWIG, GERMANY‘S VISION OF EMPIRE IN VENEZUELA: 1871–1914 45–62 (1986)
(detailing the events leading up to one such confrontation between Germany and Great Britain and the host
country of investment, Venezuela).
18. BISHOP ET AL., supra note 7, at 4.
19. DOLZER & STEVENS, supra note 1, at 8 (citing the Andean Common Market (ANCOM) as evidence that
―[m]ost Latin American countries have supported the Calvo doctrine well into the twentieth century.‖).
20. Jewswald W. Salacuse & Nicholas P. Sullivan, Do BITs Really Work?: An Evaluation of Bilateral
Investment Treaties and Their Grand Bargain, 46 HARV. INT‘L L.J. 67, 68 (2005). See also Wong, supra note 16,
at 139 (―As the global economy began to normalize following World War II, foreign capital flowed more freely
and the significance of foreign investment grew.‖).
21. BISHOP ET AL., supra note 7, at 4 (The ever-increasing recognition of property rights in human rights
declarations and state constitutions during this period made the clash between investors and governments all the
more sharply defined).
22. Id.
23. DOLZER & STEVENS, supra note 1, at 3–4.
24. BISHOP ET AL., supra note 7, at 4.
25. DOLZER & STEVENS, supra note 1, at 10–11.
26. See also id. at 11 n.35 (noting the inadequate dispute resolution procedures provided by FCNs); Franck,
supra note 6,at 1526 (―[FCNs] did not have a forum for resolving disputes.‖).
528 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
During the same period, multilateral solutions were attempted with little success.27
The Havana Charter of 1948 had the aim of establishing an international trade organization,
but it failed because of the disparate interests among developed and developing nations.28
The Economic Agreement of Bogotá of 1948 was likewise not enacted.29
Therefore, even in
the 1950s, investors‘ only support was ―an ephemeral structure consisting largely of
scattered treaty provisions, a few questionable customs, and contested general principals of
law.‖30
Before the BIT became the prevalent instrument it is today, further attempts at
multilateral treaties, though ineffective, helped frame the contours of the BITs that would
follow.
In particular, two proposals in the 1960s were significant for suggesting approaches to
balance the host country‘s sovereignty and the investor‘s need for protection. First, in 1962
the United Nations General Assembly adopted a resolution that affirmed an investor‘s
entitlement to compensation in the event of expropriation, and required good faith
observance of foreign investment agreements.31
Another U.N. resolution affirmed a host
country‘s permanent sovereignty over natural resources.32
That these resolutions were not
given formal effect is a reflection of different interpretations regarding what constituted
internationally recognized principles.33
Second, the Organization for Economic Cooperation and Development (OECD)
produced the Draft Convention on the Protection of Foreign Property in 1967.34
The Draft
was recommended to OECD member states as a model for foreign investment treaties.
Although it too failed to gain official status, ―it nevertheless played an important role in the
development and formulation of subsequent treaties negotiated by OECD member states.‖35
The first modern bilateral investment treaty had been signed between Germany and Pakistan
in 1959.36
As more European countries began to negotiate treaties with developing
countries, the consensus that had been built around the OECD Draft helped ensure that,
when member countries wrote treaties of their own, they would be working from a common
foundation.37
The success of European nations with BITs encouraged other nations to draft their
own. Sovereigns quickly found that similar treaties could enhance ―the type of asset
protection that facilitates wealth-creating cross-border capital flows, bringing net gains for
both host state and foreign investors.‖38
Like other international treaties, most BITs contain
dispute settlement provisions for inter-governmental disputes arising out of the
interpretation or application of the BIT.39
The focus of the following discussion, however,
will be those BIT provisions providing for resolution between one signatory state and
27. Salacuse & Sullivan, supra note 20, at 72 (discussing various multilateral efforts, including the 1957
International Convention for the Mutual Protection of Private Property Rights in Foreign Countries and the 1959
Abs-Shawcross Draft Convention).
28. BISHOP ET AL., supra note 7, at 4. See also Wong, supra note 16, at 140 (―Wide-ranging interests of
multiple countries . . . were ultimately too striated to reconcile.‖).
29. BISHOP ET AL., supra note 7, at 4.
30. Salacuse & Sullivan, supra note 20, at 66.
31. BISHOP ET AL., supra note 7, at 5; G.A. Res. 1803 (XVII), at 15–16, U.N. Doc. A/5217 (Dec. 14, 1962).
32. G.A. Res. 2158 (XXI), at 29, U.N. Doc. A/6518 (Nov. 25, 1966).
33. DOLZER & STEVENS, supra note 1, at 2.
34. Id. at 1–2.
35. Id. at 2.
36. Id. at 1.
37. Id. at 2.
38. Guillermo A. Alvarez & William W. Park, The New Face of Investment Arbitration: NAFTA Chapter
11, 28 YALE J. INT‘L L. 365, 366 (2003).
39. DOLZER & STEVENS, supra note 1, at 119.
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 529
citizens of the other. Such provisions are unusual insofar as they provide private parties the
right to bring a claim under an international treaty.40
By 1970, developing and developed countries had concluded a total of eighty-three
BITs.41
Eighty-six BITs were signed in the 1970s, and 211 were concluded in the 1980s.42
Beginning in the 1990s, the number of BITs increased exponentially, as emerging
economies in Eastern and Central Europe, Asia, and South America opened their markets to
attract foreign capital.43
There were slightly fewer than 400 BITs worldwide by the end of
1990,44
1857 at the end of the 1990s,45
and 2400 in place at the end of 2004.46
The BIT has
become an integral part of the legal framework for developing countries, who view them as
a highly necessary means of attracting foreign investment.47
It has been said that BITs ―are
now a touchstone of international relations.‖48
B. International Centre for the Settlement of Investment Disputes (ICSID)
1. A Forum for the Resolution of BIT Claims
For the investor-state relationship, ICSID provided a perfectly suited forum for dispute
resolution. ICSID was created by the World Bank in 1965 and made effective a year later.
ICSID provides arbitration and conciliation services for disputes between states and
investors from other states, when both states are parties to the Convention for the
International Centre for the Settlement of Investment Disputes (known as either the ICSID
Convention or the Washington Convention).49
Until the mid-1980s, ICSID usually derived
its jurisdiction from the consent given by both an investor and a state in a foreign investment
contract.50
Therefore, ICSID is not a product of BITs, at least not in the sense that only BIT
arbitration provisions can give rise to ICSID jurisdiction. Nevertheless, as BITs gained
prevalence, they have, together with multilateral treaties, become the primary source of
ICSID‘s jurisdiction.51
Although ICSID‘s caseload was slow to develop, comprising just over thirty
arbitrations in its first thirty years,52
the relevance of the institution to investor-state
40. Id.
41. UNCTAD, BILATERAL INVESTMENT TREATIES, 1959-1991 at 3, U.N. Sales No. E.92.II.A.16 (1992).
42. Id. at 4.
43. Id. at 3–4.
44. Id. at 3.
45. Course on Dispute Settlement, supra note 2.
46. UNCTAD, WORLD INVESTMENT REPORT 2005: TRANSNATIONAL CORPORATIONS AND
INTERNATIONALIZATION OF R&D at 28, U.N. Doc. UNCTAD/WIR/2005, U.N. Sales No. E.05.II.D.10 (2005).
47. UNCTAD, BILATERAL INVESTMENT TREATIES, 1959-1999, supra note 2, at 1 (―[BITs] constitute to date
the most important instrument for the international protection of foreign investment.‖). See also DOLZER &
STEVENS, supra note 1, at xi.
48. Franck, supra note 6, at 1527.
49. Course on Dispute Settlement, supra note 2, at 1.
50. Gill et al., supra note 5, at 397 & n.5.
51. Of the twenty-six arbitration proceedings registered with ICSID in 2006, twenty-one were based on
bilateral or multilateral treaties. See ICSID, 2006 Annual Report, at 5 (2006), http://icsid.world
bank.org/ICSID/FrontServlet?requestType=ICSIDPublicationsRH&actionVal=ViewAnnualReports&year=2006_E
ng (last visited Mar. 28, 2008).
52. See ICSID Official Website, List of Concluded Cases, http://icsid.worldbank.org/ICSID/
FrontServlet?requestType=CasesRH&actionVal=ListConcluded (last visited Mar. 28, 2008).
530 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
arbitration is now beyond question. Since 2000, there have been 65 ICSID tribunals.53
In
2006, ICSID administered a record 118 cases, bringing the total cases registered with ICSID
to 210.54
However, the number of arbitrations does not tell the whole story, since this is a
forum reserved for highly controversial disputes involving large sums—disputes that are far
from commonplace.55
Other forums are simply not as well suited to such disputes, leading
one practitioner to summarize as follows: ―[M]ost specialists would agree that ICSID
arbitration is an intrinsically superior mechanism in certain contexts. An ICSID award,
rendered in accordance with one of the most universally accepted international treaties, may
be expected to have an exceptionally high degree of authority.‖56
Therefore, the importance
of ICSID is apparent not so much in the sheer volume of arbitrations concluded by the
institution as in the high number of BITs pointing to ICSID arbitration.57
Furthermore, the
increasingly ubiquitous nature of BITs has required a coherent dispute resolution
framework, which ICSID provides.
Beyond ICSID‘s unique position in the world of international investment, another
reason why investors choose ICSID arbitration is the text of the BITs themselves. Dispute
resolution clauses in BITs designating ICSID arbitration in large part explain the prevalence
of ICSID arbitration in foreign investment disputes. The first BITs provided for resolution
of disputes by non-ICSID ad hoc arbitration, and not until 1968 did a BIT contain a clause
for ICSID arbitration.58
Since then, the use of such clauses has increased, and it was only a
matter of time before the number of ICSID arbitrations rose accordingly.59
As already
noted, BITs may provide for resolution of disputes before bodies other than ICSID, such as
UNCITRAL or the ICC, but ―[t]he overwhelming majority of BITs contain a reference to
ICSID.‖60
This reference is either exclusive or, more commonly, contained among other
choices for arbitration.61
The fact that ICSID arbitration is so widely used for BIT-based
claims is therefore in large part textual—a direct result of the dispute resolution clause in the
BIT designating an ICSID arbitral tribunal. The investors‘ choice of ICSID arbitration,
despite the availability of alternative forums, indicates that, with regard to investor-state
disputes, ICSID is the premier arbitration forum.
2. The Source of ICSID Jurisdiction
The dispute resolution clause in a BIT does more than designate ICSID as the forum
for arbitration; it also serves as evidence of the parties‘ consent to arbitrate BIT disputes.
Article 25(1) of the ICSID Convention establishes that ICSID jurisdiction extends to:
any legal dispute arising directly out of an investment, between a Contracting
State or any constituent subdivision or agency of the Contracting State designated
53. See id.
54. ICSID, 2006 Annual Report, supra note 51, at 5.
55. Jan Paulsson, Dispute Resolution, in ECONOMIC DEVELOPMENT, FOREIGN INVESTMENT, AND THE LAW:
ISSUES OF PRIVATE SECTOR INVOLVEMENT, FOREIGN INVESTMENT AND THE RULE OF LAW IN A NEW ERA 209, 219
(Robert Pritchard ed., 1996).
56. Id.
57. See M. SORNARAJAH, THE INTERNATIONAL LAW ON FOREIGN INVESTMENT 251 (Cambridge Univ. Press
ed., 2nd ed. 2004) (―Most dispute settlement provisions in investment treaties refer to ICSID arbitration.‖).
58. DOLZER & STEVENS, supra note 1, at 130.
59. See Gill et al., supra note 5, at 397 n.5.
60. Id. at 129.
61. Id.
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 531
to the Centre by that State, and a national of another Contracting State, which the
parties to the dispute consent in writing to submit to the Centre.62
Unless two parties ―consent in writing‖ to ICSID arbitration, a tribunal will not have
jurisdiction over a dispute between them.63
Mere reference to ICSID arbitration by a state in
a BIT does not necessarily indicate consent. For example, some dispute resolution
provisions provide for arbitration only upon subsequent agreement by the parties.64
Other
BIT provisions also stop short of giving consent, providing that a contracting state shall give
―sympathetic consideration to a request‖ for arbitration or, more forcefully, ―shall assent to
any demand [by a national of the other country].‖65
The purpose of drawing attention to
such provisions, in which parties give something less than full consent, is not to
overemphasize their importance. Indeed, in most BITs the countries give what appears to be
unambiguous consent.66
Rather, the fact that these ―consents to consent‖ exist should bring
unambiguous consent in BITs into sharp relief. For example, the UK-Sri Lanka BITs
provision beginning, ―[e]ach Contracting Party hereby consents . . .‖ evidences the parties‘
explicit consent.67
The relevant provision in the United States-Cameroon BIT reads as
follows: ―Each Party hereby consents to the submission of an investment dispute to the
International Centre for the Settlement of Investment Disputes (―ICSID‖) for settlement by
conciliation or binding arbitration.‖68
With such clear language, parties make clear to the
whole world how they intend to settle disputes. No wonder consent has been called the
―cornerstone‖ of ICSID‘s jurisdiction.69
But the parties‘ consent to the BIT solves only half of the consent equation. After all,
BITs have as their purpose the delineation of rights and responsibilities between states and
private investors. Therefore, before a tribunal can accept jurisdiction, it must be satisfied
that the investor submitting a claim to ICSID has consented to the same forum as provided
for in the BIT. The widely held view is that an investor, by submitting a claim to ICSID,
has thereby consented to its jurisdiction.70
Certainly, this result is driven in part by
practicality, since every potential investor cannot be a signatory to the BIT under which a
claim is later brought. But more than that, it is consistent with the original intention of the
ICSID Convention‘s drafters, who viewed BIT consent provisions as constituting an open
offer by a state that could subsequently be accepted by the other state‘s investors.71
The
contracting state‘s consent has been referred to as a ―generic offer‖ that is accepted upon
62. ICSID Convention, supra note 3, art. 25(1).
63. See SGS v. Philippines, ICSID Case No. ARB/02/6 paras. 26 (―The Tribunal‘s jurisdiction, if it exists,
must arise by virtue of the ICSID Convention associated with the BIT.‖), 51 (―The Philippines stressed that both
parties‘ consent is the cornerstone of ICSID jurisdiction and that such consent was not given by the Philippines
. . . .‖).
64. DOLZER & STEVENS, supra note 1, at 132 (―[F]or example, the 1979 treaty between Sweden and
Malaysia provides that: ‗In the event of a dispute . . . it shall upon the agreement by both parties to the dispute be
submitted for arbitration to [ICSID].‖) (emphasis added).
65. Id. at 133 (quoting the Netherlands-Kenya BIT and the Netherlands-Pakistan BIT, respectively).
66. Id. at 134.
67. Agreement Between the Government of the United Kingdom of Great Britain and Northern Ireland and
the Government of the Republic of Sri Lanka for the Promotion and Protection of Investment, UK-Sri Lanka, Feb.
13, 1980, 1227 U.N.T.S. 391, art. 8(1).
68. United States-Cameroon BIT, supra note 4, art. 7(3).
69. Report of the Executive Directors on the Convention on the Settlement of Investment Disputes Between
States and Nationals of Other States, para. 23, ICSID Doc. 2 (March 18, 1965) [hereinafter Directors‘ Report].
70. See Bernardo M. Cremades, Investor Protection and Legal Security in International Arbitration, 60
DISP. RESOL. J. 82, 84 (2005).
71. Directors‘ Report, supra note 69, para. 24.
532 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
making a request for arbitration.72
An underlying theme of this article is that such generic
offers by contracting states cede a large degree of authority to the arbitrator, including the
question of what types of claims an ICSID tribunal may decide.
3. The Effect of Umbrella Clauses
ICSID‘s jurisdiction derives solely from the parties‘ consent, and today this consent is
frequently found in BITs. Consequently, an investor, who for reasons of impartiality and
convenience wants ICSID to arbitrate all of the investor‘s claims, would naturally want to
describe the claims as those which the contracting state agreed to arbitrate. Suppose, for
example, that an investor enters into a concession contract with a foreign government.
Problems ensue, and the investor has both a BIT claim for unfair and inequitable treatment
and a simple breach of contract claim. Assuming the BIT confers consent, an ICSID
tribunal has jurisdiction over the BIT claim, because a country‘s consent extends to BIT
claims.73
But does ICSID have jurisdiction over the contract claim? The contract probably
has a forum-selection clause designating the host country‘s courts or tribunals as the forum
for resolution of contract disputes. Not wanting to be subject to the contentious courts of a
foreign country, the investor has every incentive to have these contract claims arbitrated
before ICSID.
While Article 25(1) gives ICSID jurisdiction over such freestanding contract claims,74
there remains the controversial issue of whether the host country has consented to arbitrate
these claims before ICSID. The investor often has difficulty arguing that the host country
consented to arbitrate purely contractual disputes before ICSID, especially since the contract
contains a forum-selection clause to the contrary. In addition, the host country will argue
that its consent to ICSID arbitration in the BIT was limited to international claims arising
out of the BIT.75
In other words, it argues that the requisite consent to arbitrate contract
claims is missing.
A new argument, based on the so-called umbrella clause found in some BITs, has
recently been added to the investor‘s arsenal. Only within the last decade have these clauses
found their way into enough BITs that ICSID tribunals have begun to grapple with their
implications.76
A typical umbrella clause provides that each contracting state agrees to
observe all of its commitments and obligations. For example, the umbrella clause in the
Switzerland-Philippine BIT provides: ―Each Contracting Party shall observe any obligation
it has assumed with regard to specific investments in its territory by investors of the other
Contracting Party.‖77
Investors argue that the host country‘s consent extends to all BIT provisions, including
the umbrella clause. The investor-state contract is one of the obligations or commitments
the host country is obligated to observe under the umbrella clause. Since ICSID‘s
jurisdiction over BIT claims is not a matter of dispute, the investor argues that ICSID has
jurisdiction over the contract claim by reason of the umbrella clause. Therefore, an ICSID
tribunal faces the question of whether the investor has, in addition to the first substantive
BIT claim, a second claim that is elevated to the level of a BIT claim by the umbrella clause.
72. See Lanco, ICSID Case No. ARB/97/6, para. 32.
73. See, e.g,. SGS v. Philippines, ICSID Case No. ARB/02/6, para. 77 (―[T]he Parties‘ intention was to
emphasize their commitment to comply with the substantive treatment obligations assumed under the BIT.‖).
74. See ICSID Convention, supra note 3, art. 25(1) and accompanying text.
75. See, e.g., SGS v. Philippines, ICSID Case No. ARB/02/6, paras. 51(b), 77.
76. See Wong, supra note 16, at 142–50 (discussing the origins and development of the umbrella clause).
77. Id. para. 34.
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 533
The importance of the umbrella clause is that it potentially tilts the debate in favor of the
investor by forcing the host country to argue that an investor-state contract is not one of the
―obligations or commitments‖ the umbrella clause encompasses. This was precisely the
issue faced in both SGS v. Philippines and SGS v. Pakistan. Both tribunals had to decide
whether the consent given in the BIT should be extended to include contract-based claims.
Before examining the SGS cases, it is necessary to review two earlier ICSID decisions,
Lanco v. Argentina and Vivendi v. Argentina. Neither case dealt with the effect of umbrella
clauses on contract claims, but their impact upon the SGS tribunals cannot be
underemphasized. Both Lanco and Vivendi examined the issue of whether a contract‘s
forum-selection clause was sufficient to dispel ICSID‘s jurisdiction over an investor‘s BIT
claim. Harmoniously, each tribunal decided that the host country‘s consent to ICSID
arbitration in the BIT was unaffected by the forum-selection clause. Vivendi went on to
wrestle with the distinction between contract-based claims and treaty-based claims,
emerging with a test that both SGS tribunals cited, though for different reasons. As will be
seen, those different reasons go a long way towards explaining the different SGS results.
III. BIT CLAIMS AND CONTRACT FORUM-SELECTION CLAUSES
A. Lanco v. Argentina
In Lanco v. Argentina, the Argentine Ministry of Economy and Public Works entered
into a concession agreement with a consortium that included Lanco International. Alleging
that Argentina breached its obligations under the Argentina-United States BIT, Lanco
submitted a request for arbitration in 1997, seeking compensation.78
A clause in the
Agreement provided for the jurisdiction of the Argentine Republic‘s courts for settlement of
disputes, and Argentina argued that the clause should preclude ICSID jurisdiction.79
In
other words, while Argentina had consented to arbitration in the BIT with the United States,
this consent was surpassed by its contractual agreement with Lanco. In the most important
part of the decision, the tribunal held Argentina to the treaty.80
It found that Argentina‘s
settlement clause in the concession agreement did not supplant consent to ICSID jurisdiction
in Article VII(4) of the BIT.81
As signatories to the BIT, the United States and Argentina
made ―a generic offer in Article VII(4) to submit to international arbitration as selected by
the investor pursuant to Article VII(3).‖82
The investor consented by submitting the dispute
for ICSID arbitration.83
Therefore, the consent of both parties required by Article 25(1) of
the ICSID Convention was present.84
The tribunal assessed Argentina‘s argument that the concession agreement nullified
this consent in light of Article 26 of the ICSID Convention:
78. Lanco, ICSID Case No. ARB/97/6, para. 2.
79. Id. para. 7.
80. Id. para. 40.
81. Id. para. 38. See United States-Argentina BIT Art. VII(4) (―Each Party hereby consents to the
submission of any investment dispute for settlement by binding arbitration in accordance with the choice specified
in the written consent of the national or company under paragraph 3 . . . .‖).
82. Lanco, ICSID Case No. ARB/97/6, para. 32.
83. Id. para. 33.
84. Id. para. 40. See ICSID Convention, supra note 3, art. 25(1).
534 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
Consent of the parties to arbitration under this Convention shall, unless otherwise
stated, be deemed consent to such arbitration to the exclusion of any other
remedy. A Contracting State may require the exhaustion of local administrative
or judicial remedies as a condition of its consent to arbitration under this
Convention.85
The second sentence of Article 26 indicates that a stipulation requiring the exhaustion
of local remedies as a condition to consent, at the very most, dilutes the exclusivity of ICSID
as a source of remedy. But it cannot replace the consent given by the contracting state.86
All of this assumes that there is such a stipulation in the first place. Significantly, the court
found no such stipulation in the BIT and was highly doubtful that one existed in either the
Agreement or the national law of Argentina.87
Not only did the Agreement‘s forumselection
clause fail to dislodge Argentina‘s consent to arbitration in the BIT, but the
tribunal made clear that this would have been the result even if Argentina had qualified its
consent in the BIT by requiring the investor to exhaust local remedies.88
B. Vivendi v. Argentina
The award in Vivendi v. Argentina involved a similar question about the relationship
between a BIT claim and a forum-selection clause in a concession contract. Compagnie
Générale des Eaux (CGE) (the predecessor in interest to Vivendi Universal) and an affiliate,
Compañia de Aguas del Aconquija, S.A. (CAA), entered into a water and sewage contract
with Tucumán, a province of Argentina.89
Shortly after beginning performance of the
concession contract, Argentina took steps that made it ―impossible . . . to operate the
concession in a commercially reasonable manner.‖90
CGE filed for ICSID arbitration,
alleging that Tucumán‘s actions and failures to act were attributable to the federal
government of Argentina, resulting in a breach of the 1991 Argentine-French BIT.91
Argentina objected to the ICSID tribunal‘s jurisdiction, arguing that the concession
contract‘s forum-selection clause, which pointed to the contentious administrative Tucumán
courts, was exclusive. As such, the clause precluded an ICSID tribunal from deciding the
dispute. Relying on Lanco, the tribunal held that it had jurisdiction over CGE‘s BIT claims.
The forum selection-clause ―did not and could not constitute a waiver‖ by CGE of its right
to bring a BIT claim before an ICSID tribunal.92
By signing the BIT, Argentina had
consented to ICSID arbitration, and a concession contract‘s forum designation could do
nothing to alter this consent. Significantly, the tribunal focused on the treaty-based nature of
the claim: ―Claimant‘s claims ‗are not subject to the jurisdiction of the contentious
85. ICSID Convention, art. 26, quoted in Lanco, ICSID Case No. ARB/97/6, para. 35.
86. Lanco, ICSID Case No. ARB/97/6, para. 38.
87. Id. para. 32 (The Argentina-United States treaty was the first in which Argentina submitted to
international arbitration without reservations).
88. Id. para. 38.
89. Vivendi, ICSID Case No. ARB/97/3, para. 11.
90. Stanimir A. Alexandrov, The Vivendi Annulment Decision and the Lessons for Future ICSID
Arbitrations, in ANNULMENT OF ICSID AWARDS 97, 98–99 (Emmanuel Gaillard et. al eds., 2004).
91. Vivendi, ICSID Case No. ARB/97/3, para. 11. In addition to the attribution argument, CGE alleged that
the Republic of Argentina breached the BIT in its own capacity. Id. para. 16. The tribunal found that Argentina
acted appropriately and dismissed these claims, a result endorsed by the Annulment Committee. Id. paras. 90–91.
92. Vivendi v. Argentina, Award, ICSID No. ARB/97/3 (1997) para. 53 [hereinafter Vivendi Award].
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 535
administrative tribunals of Tucumán, if only because, ex hypothesi, these claims are not
based on the Concession Contract but allege a cause of action under the BIT.‘‖93
The tribunal, however, stopped short of deciding the merits of the investor‘s BIT
claim. It found that examining the merits of CGE‘s BIT claim required a determination of
whether Tucumán had breached the contract, and the forum-selection clause prohibited such
a ―detailed interpretation.‖94
An ICSID tribunal could hear the merits of the claim only after
the claimant had first submitted the dispute to the courts of Tucumán—and then only if the
local courts had ―procedurally or substantively‖ denied CGE‘s rights.95
Thus, while the
contract‘s forum-selection clause did not prevent ICSID jurisdiction, the tribunal seized on
it as a basis for refusing to decide the merits of the claim.
The claimants sought an annulment of the award by special committee. Their
argument was that the tribunal, by refusing to resolve the merits of the dispute, manifestly
exceeded its powers—one of three grounds for annulment provided for in Article 52 of the
ICSID Convention.96
The committee proceeded by examining the tribunal‘s findings on the
issues of both jurisdiction and the merits.97
It reaffirmed the tribunal‘s holding that a
contract could not act as a barrier to ICSID jurisdiction over a BIT claim. Of primary
importance was the committee‘s reaffirmation of the Lanco doctrine that a contract between
a sovereign and an investor did not nullify the sovereign‘s consent to arbitrate before ICSID
claims arising under the BIT.98
But with jurisdiction comes responsibility. If the tribunal
had jurisdiction over the dispute, it followed that anything less than a decision on the merits
was an abrogation of the tribunal‘s responsibility rising to the level of a manifest excess of
powers. ―In the Committee‘s view, it is not open to an ICSID tribunal having jurisdiction
under a BIT in respect of a claim based upon a substantive provision of that BIT, to dismiss
the claim on the ground that it could or should have been dealt with by a national court.‖99
By declining to decide the merits, the tribunal had made it impossible for the claimants
to later bring a BIT claim before ICSID.100
The annulment committee emphasized the close
relationship between finding jurisdiction and deciding the merits. Just as a forum-selection
clause is insufficient to remove a BIT claim from ICSID jurisdiction, a contract claim does
93. Vivendi, ICSID Case No. ARB/97/3, para. 14(d).
94. Vivendi Award, para. 79.
95. Vivendi Award, para. 78.
96. See ICSID Convention, supra note 3, art. 52 (The claimants also argued that if the tribunal had decided
the merits of the dispute, as opposed to dismissing the case, it had done so without stating reasons as provided in
Article 52. This was rejected by the committee as grounds for annulment.).
97. The Committee reexamined the jurisdictional issue in response to Argentina‘s request for a conditional
annulment of the jurisdictional aspect of the award. Vivendi, ICSID Case No. ARB/97/3, para. 67.
98. Vivendi, ICSID Case No. ARB/97/3, paras. 76–78.
99. Id. para. 102.
100. The committee found particularly relevant Article 8(2) the France-Argentina BIT, which contained a socalled
―fork in the road‖ provision. Article 8(2) says that ―[o]nce an investor has submitted the dispute to the
courts of the Contracting Party concerned or to international arbitration, the choice of one or the other of those
procedures is final.‖ The tribunal was aware that requiring CGE to submit the breach of contract issue to the
contentious administrative courts of Tucumán potentially closed the door on ICSID‘s jurisdiction. As such, the
tribunal attempted to diminish the importance of such a drastic step by reading the ―fork in the road‖ provision
narrowly to apply only to pure BIT claims. Since the claimants would be bringing a contract claim to the local
courts (as opposed to a BIT claim), the tribunal was confident that art. 8(2) did not foreclose the possibility of later
bringing a BIT claim before ICSID. That is, a fork in the road was never reached, because the claimant‘s
submission to the local courts was contractual in nature. The committee disagreed and imposed a broader reading
of art. 8(2), with an important result: because 8(2) is not limited exclusively to claims alleging a breach of the
BIT, requiring claimants to bring their non-BIT claims to local courts would have constituted an absolute waiver
of their right to ICSID arbitration. Therefore, the tribunal‘s refusal to decide the merits had harsher consequences
for the claimants than the tribunal was willing to admit. Id. para. 36.
536 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
not eclipse a BIT claim just because there is overlap between a BIT breach and a contract
breach. As a result, the tribunal‘s finding that CGE‘s BIT claims were so ―intertwined‖ with
the contract claims as to require an analysis of the contract itself was no reason to refuse to
decide the merits. It was not irrelevant that the contract may have been breached. But the
BIT claims ―set an independent standard. A state may breach a treaty without breaching a
contract, and vice versa, and this is certainly true of these provisions of the BIT.‖101
Deciding whether there has been a breach of the BIT is a question of international law,
resolved by reference to the obligations each BIT country assumes. By contrast, in order to
decide a contract question, reference to the local law of the country is required. The
committee allowed that ―compliance with internal law is relevant to the question of
international responsibility. But this is because the rule of international law makes it
relevant, e.g., by incorporating the standard of compliance with internal law as the
applicable international standard or as an aspect of it.‖102
The law of Tucumán was relevant
in determining whether Tucumán breached the contract. This in turn indicates whether
Argentina breached a provision of the BIT. The overlap between the two issues should not
absolve an ICSID tribunal of having to decide the merits of the BIT claim. As the
committee made clear, whether ICSID jurisdiction exists depends on the essence of the
claim:
[W]here ―the fundamental basis of the claim‖ is a treaty laying down an
independent standard by which the conduct of the parties is to be judged, the
existence of an exclusive jurisdiction clause in a contract between the claimant
and the respondent state or one of its subdivisions cannot operate as a bar to the
application of the treaty standard.103
A few points regarding this fundamental basis test will be particularly relevant to the
discussion that follows. First, there was no objection in Vivendi to the investors‘ description
of the claim as treaty-based. In fact, the committee‘s conclusion affirmed its status as a BIT
claim by emphasizing its separateness from a factually related contract claim.104
This leaves
open the question of what a tribunal must do if it accepts jurisdiction over a claim that is not
fundamentally treaty-based. If a tribunal accepts jurisdiction and decides the merits of a
contract claim, this could be seen as violating the forum-selection clause of the contract with
no apparent justification found in Vivendi.
The justification lacking in Vivendi, but present in the SGS cases and others, is an
umbrella clause. An ICSID tribunal cannot accept jurisdiction over a contract claim based
on Vivendi‘s fundamental basis test, but an umbrella clause changes the analysis. Similarly,
a tribunal that follows Vivendi and takes the opposite stance—giving effect to the forumselection
clause based solely on the contractual nature of the claim—may be ignoring an
important variable. The presence of an umbrella clause raises difficult questions about its
application, and while tribunals have reached divergent answers, one of its clear effects is to
alter the stand-alone status of Vivendi. In short, the clause is a fulcrum with the potential to
alter Vivendi‘s ―fundamental basis‖ categories.
Finally, the Vivendi annulment committee fashioned the fundamental basis test only
after affirming the tribunal‘s holding regarding jurisdiction. Looking to Lanco, the tribunal
had found that it had jurisdiction over the BIT claims because the parties so consented. The
101. Id. para. 95.
102. Id. para. 70.
103. Id. para. 101.
104. Vivendi, ICSID Case No. ARB/97/3, para. 96.
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 537
annulment committee agreed.105
Therefore the jurisdictional question was separate from the
merits of the case. The committee‘s basis for annulment was the tribunal‘s refusal to decide
the merits of the BIT claim on the grounds of its close connection to the contract claim, and
it was in this connection that the committee articulated its oft-cited fundamental basis test.
Therefore any use of the fundamental basis test by a tribunal prior to accepting jurisdiction
would be premature, or at the very least a misapplication of Vivendi. Yet as the SGS v.
Pakistan tribunal‘s decision demonstrates, test is susceptible to use for purposes other than
those intended by the Vivendi committee.
IV. THE SGS CASES: OF CONTRACT CLAIMS AND FORUM-SELECTION
CLAUSES
A. Declining Jurisdiction over Contract Claims: The Approach in SGS v. Pakistan
1. Background
The dispute between SGS Société Générale de Surveillance S.A. and the Islamic
Republic of Pakistan arose from a Pre-Shipment Inspection (PSI) Agreement the two parties
signed in 1994.106
Approximately two years later, Pakistan notified SGS that the contract
would be terminated. The procedural history that followed was more involved and
prolonged than the average pre-ICSID arbitration maneuvering. After communications
between the parties failed to reach a resolution, SGS filed suit against Pakistan in
Switzerland. Pakistan successfully argued that in the PSI Agreement the parties had chosen
to arbitrate any disputes arising out of the contract before an arbitral tribunal in Pakistan.107
On this ground, the Swiss Court of First Instance rejected SGS‘ claim, and SGS‘ subsequent
appeals were unavailing.108
Meanwhile, Pakistan had instituted arbitral proceedings in
Pakistan, invoking the arbitration provided for in the PSI Agreement.109
SGS objected to
the PSI Agreement arbitration but filed counter-claims against Pakistan for alleged breaches
of the Agreement.110
Six months later, SGS notified Pakistan that it was seeking resolution
of the dispute according to Article 9(2) of the Swiss-Pakistan BIT, which provided for
ICSID arbitration.111
In its request for ICSID arbitration, SGS alleged several claims under the BIT,
including failure to ensure ―fair and equitable treatment‖ of SGS‘ investment and taking
measures of expropriation without providing effective and adequate compensation.
Furthermore, in what would become the decision‘s flashpoint issue, SGS asserted that
Pakistan violated Article 11 of the BIT by failing to guarantee observance of its contractual
commitments. Article 11 reads: ―Either Contracting Party shall constantly guarantee the
105. Id. para. 77.
106. SGS v. Pakistan, ICSID Case No. ARB/01/13, para. 11.
107. Id. paras. 22–25.
108. Id. paras. 23–25. The higher Swiss courts upheld the result on the ground that Pakistan, because it
enjoyed sovereign immunity, could not be sued in Swiss courts.
109. Id. para. 26.
110. Id. para. 27.
111. Id. para. 32.
538 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
observance of the commitments it has entered into with respect to the investments of the
investors of the other Contracting Party.‖112
By this umbrella clause, SGS sought to bring its contractual claims within the ambit of
the BIT, so that the ICSID tribunal would decide not just BIT claims like expropriation and
inequitable treatment, but purely contractual claims as well. Pakistan objected that the
ICSID tribunal could not investigate these Article 11 claims until the Pakistani arbitral
tribunal made a determination as to whether there had been a contract breach in the first
place.113
Moreover, the parties had chosen as their method of dispute resolution arbitration
in Pakistan, and accepting SGS‘ view would ―eviscerate the parties‘ specific arbitration
agreement in this case.‖114
For its part, SGS insisted that Article 11 ―elevated‖ otherwise
purely contractual claims into the favored realm of BIT claims over which ICSID had
jurisdiction.115
The tribunal ultimately sided with Pakistan and decided it had no jurisdiction over
contractual claims.116
But en route to this result, it decided conclusively that it had
jurisdiction over SGS‘ BIT claims, purporting to follow Lanco and Vivendi. An
examination of the tribunal‘s reasoning regarding its jurisdiction over BIT claims helps in
understanding why the tribunal found it lacked jurisdiction over contract claims.
2. Analysis
As a preliminary matter, the Pakistan tribunal had to decide whether to allow a
claimant to characterize claims as it sees fit, without the tribunal inquiring too closely into
the ―real‖ nature of the claim. Pakistan argued forcefully that SGS was characterizing as
BIT claims what in substance were mere contract claims. For four years preceding the
request for ICSID arbitration, SGS had never raised a BIT issue, content to frame its claims
as contract-based. Mere re-labeling as a BIT claim before ICSID was insufficient to mask
the claim‘s true contractual nature, Pakistan argued.117
Pakistan drew upon Vivendi‘s
fundamental basis test, arguing that ―[i]n a case where the essential basis of a claim brought
before an international tribunal is a breach of contract, the tribunal will give effect to any
valid choice of forum clause in the contract.‖118
SGS responded accordingly, citing Vivendi
for the proposition that a BIT-labeled claim is per se a treaty-based claim.119
Unfortunately, the tribunal did not acknowledge this misuse of the Vivendi standard,
which, as noted above, had as its purpose the determination of whether a tribunal must
decide a claim‘s merits.120
Passing for the time being on the question of the claims‘
fundamental basis, the tribunal simply endorsed SGS‘ pragmatic argument that the benefit of
the doubt should go to the claimant asserting a BIT violation; not until the merits phase
should the tribunal probe the veracity of the claim‘s BIT characterization.121
But if the
tribunal is untroubled by the prospect of separating BIT claims from contract claims after
112. Switzerland-Pakistan BIT, art. 11.
113. See SGS v. Pakistan, ICSID Case No. ARB/01/13, para. 48.
114. Id. para. 55.
115. Id. para. 54.
116. Id. para. 173.
117. Id. para. 62.
118. Id. para. 44.
119. SGS v. Pakistan, ICSID Case No. ARB/01/13, para. 83.
120. See supra, pp. 14–15.
121. SGS v. Pakistan, ICSID Case No. ARB/01/13, para. 145 (―At this stage of the proceedings, the Tribunal
has, as a practical matter, a limited ability to scrutinize the claims as formulated by the Claimant . . . . ‖).
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 539
accepting jurisdiction, it makes little sense to deny itself jurisdiction based on the
contractual nature of a claim under any circumstances. But as the rest of its decision
demonstrates, the tribunal saw no incongruity between this generous standard and a narrow
umbrella clause interpretation.
Next, the tribunal confronted the same question that faced the Lanco and Vivendi
tribunals: did it have jurisdiction over the BIT claims? Again, Pakistan relied on Vivendi‘s
fundamental basis test to challenge the tribunal‘s jurisdiction.122
As before, the tribunal did
not attempt to clarify as non-jurisdictional the portions of the Vivendi opinion cited by
Pakistan. This was the time to look to Lanco and Vivendi to see how those tribunals
assessed the jurisdiction question. Instead, the tribunal tackled the issue as it was posed,
asking whether overlap between contract and treaty claims was sufficient to nullify ICSID
jurisdiction. BIT and contract claims each present distinct questions to be resolved by
different laws, the tribunal decided, citing the Vivendi annulment committee‘s discussion of
the separateness of a treaty breach and a contract breach.123
The tribunal concluded that it
had jurisdiction over those claims whose fundamental basis was in the BIT.124
On its way to this holding, the Tribunal examined Article 9 of the BIT, in which the
parties consented to ICSID arbitration. This clause gave little direction as to the types of
claims the parties consented to arbitrate, leaving the tribunal to declare that ―if Article 9
relates to any dispute at all . . . it must comprehend disputes constituted by claimed
violations of BIT provisions establishing substantive standards of treatment . . . .‖125
The
tribunal acknowledged that the contract‘s forum-selection clause was insufficient to divest
the tribunal of jurisdiction,126
but this seems more like an afterthought than an endorsement
of Lanco and Vivendi‘s analysis.127
The tribunal accepted jurisdiction over the BIT claims
because they were fundamentally treaty claims and because there was no good reason to
decline jurisdiction.
By accepting jurisdiction over BIT claims, the Pakistan tribunal reached the right
result, but for the wrong reasons. The tribunal would have done better to examine Vivendi‘s
discussion of the jurisdictional issue, which relied heavily on Lanco. Instead, it lost sight of
Lanco‘s focus on consent between sovereigns as the basis of BIT arbitration agreements.
The tribunal went on to hold that it lacked jurisdiction over contract claims, which is
understandable, if only because such a holding resides so comfortably alongside the
application of the Vivendi test to the issue of BIT claims. If the entire universe of BIT
claims is limited to those claims whose ―fundamental basis‖ is in the BIT, then it is little
surprise that a contract claim does not meet the Pakistan tribunal‘s standard because its
fundamental basis will always be contractual. Of course, the Pakistan tribunal confronts the
question of whether the umbrella clause transforms what is fundamentally a contract claim
122. Id. paras. 62–63.
123. Id. paras. 147–48.
124. Id. para. 155.
125. Id. para. 150.
126. SGS v. Pakistan, ICSID Case No. ARB/01/13, para. 154.
127. Id. para. 152 (noting that the parties did not give ICSID exclusive jurisdiction over BIT claims). Cf.
Lanco, ICSID Case No. ARB/97/6, para. 36 (―[W]hen the parties give their consent to ICSID arbitration, they lose
their right to seek to settle the dispute in any other forum . . . .‖); id. para. 38 ([A dispute resolution clause in
contract] does not replace any consent but instead dilutes the presumption as to the exclusivity of ICSID
jurisdiction.‖). The tribunal‘s analysis casts doubt as to whether its BIT jurisdiction would be exclusive if the
facts were slightly different. SGS v. Pakistan, ICSID Case No. ARB/01/13, para. 153 (reasoning that parties
contracting prior to BIT‘s entry into force do not intend dispute resolution clause to relate to BIT claims); id. para.
154 (noting that no claims based on the BIT have been brought pursuant to contract‘s dispute resolution
provision).
540 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
into the BIT claim. But against the backdrop of its BIT claim reasoning, one might
speculate that the tribunal was approaching the problem in reverse, that is, beginning with
the general principal that a contract claim can never become a BIT claim. The reasons
given in support of its holding on the issue of contract claims do little to dispel such a
conclusion.
The first reason given by the tribunal for reading Article 11 narrowly is that a broad
interpretation would render the umbrella clause ―susceptible to almost indefinite
expansion.‖128
The result would be to make an international obligation out of all ―the
municipal legislative or administrative or other unilateral measures of a Contracting Party.‖
There are two problems with this argument. The first is based on the tribunal‘s role, which
is limited to deciding a dispute between the parties based on the text of the BIT presented.
If susceptibility to indefinite expansion is the primary reason for the tribunal‘s holding, it is
inadequate. Such reasoning has its basis not in the application of Article 11 to the contract
claims at issue, but in possible reprecussions for tribunals interpreting similarly worded
BITs. Second, as the SGS v. Philippines tribunal later pointed out, the umbrella clause
would not apply to any and all legal obligations, but only to those ―assumed vis-à-vis the
specific investment . . . .‖129
For example, a unilateral obligation promulgated by a
contracting state as part of its national law would not be encompassed by an umbrella
clause. Thus, to say that an umbrella clause would have limitless application is an untenable
basis for denying a readily identifiable contract claim its coverage.
The tribunal advanced a second argument, focusing on the principle in international
law that ―a violation of a contract entered into by a State with an investor of another State, is
not, by itself, a violation of international law.‖130
Cited approvingly in Vivendi, this
principle loses force when an umbrella clause is involved. A clause like Article 11 of the
Swiss BIT is exactly the kind of mechanism that would elevate a contract violation to the
level of an international violation. At the very least, ―[t]he question is essentially one of
interpretation, and does not seem to be determined by any presumption.‖131
It is incomplete
to assert that a contract violation is not ―by itself, a violation of international law‖ when at
issue is a clause whose operation would render the contract violation a violation of
international law.
Finally, the Pakistan tribunal voiced its concern that a broad interpretation of an
umbrella clause would override forum-selection clauses in investor-state conflicts.132
Of all
the reasons given for declining jurisdiction, this deserves the most attention, because it
points to a very real problem that arises once jurisdiction is accepted. But as the Philippines
tribunal demonstrated, acceptance of a contract claim based on the umbrella clause does not
require resolution of the contractual dispute. Pakistan, however, assumed that it did.
The tribunal cannot be faulted for its concern that a contract‘s forum-selection clause
would bow to the sweeping language of the umbrella clause. After all, this is the tension at
the heart of the SGS cases. But in part because of its analysis of the BIT issue, in which it
erroneously applied the Vivendi test, the tribunal was ill-equipped to confront the issue of
contract claims. Had it instead consulted the Lanco standard with regard to the BIT issue,
the tribunal might have been more likely to grant the appropriate breadth to the SwitzerlandPakistan
BIT‘s umbrella clause when deciding whether to afford contract claims BIT
protection.
128. Id. para. 166.
129. SGS v. Philippines, ICSID Case No. ARB/02/6, para. 121.
130. SGS v. Pakistan, ICSID Case No. ARB/01/13, para. 167.
131. SGS v. Philippines, ICSID Case No. ARB/02/6, para. 122.
132. SGS v. Pakistan, ICSID Case No. ARB/01/13, para. 168.
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 541
The position taken by SGS v. Pakistan has support among ICSID tribunals. Joy
Mining v. Egypt found that an umbrella clause not prominently inserted into a BIT cannot
operate to transform a contract claim into a treaty claim, unless there exists ―a clear violation
of the Treaty rights and obligations or a violation of contract rights of such a magnitude as
to trigger the Treaty protection . . . .‖133
Since all of the claimant‘s causes of action in that
case were contractual in nature and did not relate to an investment, there were no treaty
claims, and the tribunal found that it lacked jurisdiction.134
Apparently this rendered the
forum-selection clause in the contract decisive with respect to contract claims.135
The
tribunal cited Vivendi, emphasizing that the ―essential basis‖ of the claim was contractual.136
This result was further compelled, according to the tribunal, by the ―missing link‖ between
the contract and the treaty able to support ICSID jurisdiction over the contract claims.137
It
seems that the umbrella clause—absent in Vivendi but present in Joy Mining—was an
insufficient link to support jurisdiction in the eyes of the tribunal. 138
The tribunal in CMS v. Argentina, analyzing the Argentina-United States BIT, allowed
that ―not all contract breaches result in breaches of the Treaty,‖ even when there is an
umbrella clause present.139
While ―purely commercial aspects of a contract‖ do not receive
treaty protection, the umbrella clause operates to elevate a contract dispute to a treaty claim
when there is ―significant interference by governments or public agencies with the rights of
the investor.‖140
The tribunal concluded that the state measures complained of by CMS were
not commercial but instead ―resulted in the violation of the standards of protection under the
Treaty.‖141
This result is consistent with Pakistan and Joy Mining in that the umbrella
clause itself is insufficient to elevate a contract claim. Yet the CMS tribunal was willing to
give meaning to the umbrella clause in the presence of ―significant interference‖ by the host
government. Here is a circumstance in which, contrary to Pakistan, an ICSID tribunal has
employed the umbrella clause despite a contract‘s forum-selection clause.142
133. Joy Mining Mach. Ltd. v. Egypt, Award on Jurisdiction, ICSID No. ARB/03/11, para. 81 (2004).
134. Id. para 63. The tribunal‘s lack of subject matter jurisdiction was based on its finding that there was no
investment within the meaning of Article 25(1) of the ICSID Convention. This appears to make the umbrella
clause discussion dicta. Alternatively, Joy Mining can be distinguished from the SGS cases on the basis that the
contractual claim at issue was not related to an investment—a factor that, because it destroyed subject matter
jurisdiction, persuaded the court to give effect to the forum-selection clause. Id. para. 89. See Sempra Energy
International v. Argentine Republic, Decision on Objections to Jurisdiction, ICSID No. ARB/01/16, para. 95
(2005) (―A claim can have a purely contractual origin and refer to a right that does not qualify as an investment, in
which case there will be no jurisdiction, as was the case in Joy; . . . or, as is more frequently the case, [a claim can]
originate in violation of a contractual obligation that at the same time amounts to a violation of the guarantees of
the treaty.‖).
135. Id. para. 89.
136. Id. para. 90.
137. Id. para 81.
138. See Wong, supra note 16, at 160 n.162 (criticizing the reasoning in Joy Mining as ―incomprehensible
and quite far from amounting to a clarification‖) (citing CHRISTOPH SCHREUER, INVESTMENT TREATY
ARBITRATION AND JURISDICTION OVER CONTRACT CLAIMS—THE VIVENDI CASE CONSIDERED, IN INTERNATIONAL
INVESTMENT LAW AND ARBITRATION 280, 301 [Todd Weiler ed., 2005]).
139. CMS Gas Transmission Co. v. Argentine Republic, Award, ICSID No. ARB/01/8, para. 299 (2005).
140. Id. para. 299.
141. Id. paras. 301–03.
142. It seems that the Pakistan, Joy, and CMS tribunals are employing the same general rule—that the
umbrella clause does not elevate a contract claim—while differing in their articulation and application of a
common exception. Pakistan (1) insists on a narrow reading, (2) allows that in extreme circumstances the
contract claim could be elevated to a treaty claim, and (3) finds that no such circumstances apply to its facts. Joy
(1) adopts a narrow reading, (2) looks for a clear violation of treaty rights to link the contract claim and the treaty,
and (3) finds no such violation on its facts. CMS (1) adopts a relatively narrow reading, (2) requires the presence
of substantial state interference before giving effect to the umbrella clause, and (3) finds sufficient interference to
542 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
In El Paso v. Argentina, the tribunal followed Pakistan‘s narrow reading of the
umbrella clause in interpreting the United States-Argentina BIT at issue. While stating that
―[t]he arguments put forward by the [Pakistan] Tribunal are . . . more than conclusive,‖143
the tribunal underscored various points of agreement with Pakistan and even buttressed the
Pakistan reasoning with arguments of its own. A narrow reading of the umbrella clause was
necessary in the tribunal‘s view to prevent a state‘s minor commitments from being
transposed to international treaty obligations.144
The tribunal expressed doubt that investors
would exercise restraint in invoking the umbrella clause: ―[t]he responsibility for showing
appropriate restraint rests rather in the hands of the ICSID tribunals.‖145
The tribunal
understandably wanted to retain discretion as to whether it would decide contract claims. In
this connection, it opted for Pakistan‘s bright-line exclusion of contract claims over the
broad reading of the umbrella clause advanced in SGS v. Philippines.146
B. Accepting Jurisdiction Over Contract Claims: The Approach of SGS v. Philippines
1. Background
The contractual relationship giving rise to the dispute between SGS and the
Philippines was similar to that in SGS v. Pakistan.147
In 1991, the Philippines entered a
contract with SGS in which the latter would provide pre-shipment inspection services of the
Philippines‘ imports in the country of supply, including verification of the imports‘ quality,
quantity, and price.148
As part of the Philippines‘ comprehensive import supervision service
(CISS), the contract was referred to as a CISS Agreement. From 1986 to 1998, new
agreements were entered into, and after 1998, the CISS Agreement was extended twice.149
In 2000, SGS‘ services under the Agreement were discontinued.150
SGS submitted claims to
the Philippines for money unpaid on the contract, totaling approximately US $140 million
plus interest.151
When attempts for amicable settlement failed, SGS submitted a request for arbitration
to ICSID, contending that the Philippines was in breach of Articles IV(1), IV(2), VI(1), and
X(2) of the Switzerland-Philippines BIT.152
The Article IV claim alleged failure to accord
SGS fair and equitable treatment, the Article VI claim concerned expropriation, and the
Article X claim was the umbrella clause requiring the host country to observe its
support implementation of the umbrella clause on its facts. Assuming this is indeed the approach intended by the
tribunals, it is not clear why a provision as sweeping as the umbrella clause should be caged inside a narrow
interpretation and released only upon the happening of an exception deemed sufficient by the tribunal. It would
seem less cumbersome to give the umbrella clause its plain broad effect and only then determine whether state
action warrants review of the contract claim by an ICSID tribunal—the approach adopted in SGS v. Philippines.
See also Shany, supra note 15 (highlighting SGS v. Pakistan and SGS v. Philippines as examples of two
competing analytical methodologies, disintegrationism and integrationism, respectively).
143. El Paso Energy International Company v. Argentine Republic, Decision on Jurisdiction, ICSID No.
ARB/3/15, para. 71 (2006).
144. Id. para. 76.
145. Id. para. 82.
146. Id. para. 76.
147. See SGS v. Philippines, ICSID No. ARB/02/6, paras. 12–13.
148. See id.
149. Id. paras. 13–14.
150. Id. para. 14.
151. Id. para. 15.
152. Id. para. 16.
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 543
commitments.153
The Philippines objected to the ICSID tribunal‘s jurisdiction. The dispute,
the Philippines argued, was ―purely contractual in character‖ and therefore subject to the
forum selection clause in the CISS agreement, which required contract disputes to be settled
in the courts of the Philippines.154
The Article IV and VI claims were contractual and not
BIT claims at all. As for the Article X claim, the Philippines relied on SGS v. Pakistan,
which had found umbrella clauses insufficient to transform a claim from contract to the
realm of BIT.
SGS countered that ICSID‘s jurisdiction was found in Article 25(1) of the ICSID
Convention, which provided for jurisdiction over a legal dispute arising directly out of an
investment when there was consent by the parties in writing.155
The required consent in
writing was present in the Switzerland-Philippines BIT,156
and SGS argued that the tribunal
had jurisdiction over contract claims based inter alia on the umbrella clause.157
Article X(2)
reads: ―Each Contracting Party shall observe any obligation it has assumed with regard to
specific investments in its territory by investors of the other Contracting Party.‖158
The tribunal‘s first task was to evaluate the strength of SGS‘ argument for a broad
umbrella clause interpretation, which SGS summarized as follows: ―[I]t [is] abundantly clear
that a provision such as Art. X(2) of the BIT is designed to extent [sic] the BIT‘s protection
to cover breaches of contracts between the host State and the investor.‖159
The tribunal
acknowledged the SGS v. Pakistan tribunal‘s conclusion with regard to the umbrella clause.
Pakistan said that such a clause should never automatically elevate a contract claim to the
level of an international breach of law, and only under ―exceptional circumstances‖ would it
provide an appropriate basis for accepting jurisdiction over a contract claim.160
The SGS v.
Philippines tribunal proceeded to examine Pakistan‘s conclusions,161
but only after reaching
its own ―provisional conclusion—that Article X(2) means what it says . . . .‖162
The tribunal advanced a number of arguments in support of bringing contract claims
within the scope of the umbrella clause. First, the text of the provision favors a broad
interpretation. It is phrased in mandatory language, using the word ―shall‖ just as in other
substantive BIT provisions imposing obligations on the parties.163
Also, the tribunal
decided that the phrase ―any obligation‖ is wide enough to include obligations arising under
national law with regard to an investor-state contract.164
―Interpreting the actual text of
Article X(2),‖ the tribunal reasoned,
[i]t would appear to say, and to say clearly, that each Contracting Party shall
observe any legal obligation it has assumed, or will in the future assume with
regard to specific investments covered by the BIT. Article X(2) was adopted
153. See SGS v. Philippines, ICSID No. ARB/02/6 paras. 113, 160.
154. Id. para. 17; para. 22 (―Article 12 of the CISS Agreement provided that: ‗. . . All actions concerning
disputes in connection with the obligations of either party to this Agreement shall be filed at the Regional Trial
Courts of Makati or Manila.‘‖).
155. SGS v. Philippines, ICSID No. ARB/02/6 para. 16. See ICSID Convention, supra note 3, art. 25(1).
156. SGS v. Philippines, ICSID No. ARB/02/6, para. 46.
157. Id. paras. 63–68.
158. Id. para. 34.
159. Id. para. 86. See also para. 65 (reviewing SGS‘ argument that ―[t]he effect of an ‗umbrella clause‘ . . .
is to elevate a breach of contract claim to a treaty claim under international law.‖).
160. Id. para. 96(b).
161. Id. paras. 120–126.
162. SGS v. Philippines, ICSID No. ARB/02/6, para. 119.
163. Id. para. 115.
164. Id. para. 115.
544 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
within the framework of the BIT, and has to be construed as intended to be
effective within that framework.165
Second, any uncertainty regarding the scope of Article X(2) should be resolved in
favor of protecting investment, since this is the overriding purpose of the BIT.166
Moreover,
―it seems entirely consistent with the object and purpose of the BIT to hold that [binding
obligations or commitments such as investor-state contracts] are incorporated and brought
within the framework of the BIT by Article(2).‖167
Finally, had the parties intended to limit
the umbrella clause to obligations arising under ―other international law instruments,‖
Switzerland and the Philippines were free to stipulate such a restriction in the BIT, but they
did not do so.168
The tribunal then examined the SGS v. Pakistan tribunal‘s conclusions, some of which
are reviewed in the previous section of this article.169
Two of the Pakistan tribunal‘s
arguments were singled out by SGS v. Philippines and will be considered here. First, SGS v.
Pakistan had found significant the location of the clause at the end of the BIT.170
Whereas
other BITs containing such clauses placed them with the treaty‘s other substantive
provisions at the beginning of the treaty, the Switzerland-Philippines BIT placed the
umbrella clause at the end.171
This was a factor worthy of consideration, the Philippines
tribunal agreed. But unlike Pakistan, the tribunal did not find the location of the clause
decisive. It reasoned that other Philippine BITs, for example, contained an umbrella clause
earlier in the treaty. The tribunal found it ―difficult to accept that the same language in other
Philippines BITs is legally operative, but that it is legally inoperative in the SwissPhilippines
BIT merely because of its location.‖172
A second Pakistan argument was particularly relevant to the rest of the Philippines
decision. According to Pakistan, a broad interpretation would have the detrimental effect of
overriding the dispute settlement clauses in investor-state contracts. If an ICSID tribunal
finds that a BIT party consented to ICSID jurisdiction over a contract claim by operation of
the umbrella clause, the host country will argue that the contract‘s forum-selection clause
should be unaffected. After all, as the Philippines argued, all indications are that an
investor-state contract is the result of bidding by various companies, followed by an armslength
agreement.173
Why should the giving of consent to arbitrate disputes with an investor
override a valid contractual clause providing for dispute settlement in the host country? The
tribunal found this to be a valid concern, but proceeded to explain why a broad
165. Id.
166. Id. para. 116. This investor-friendly approach was criticized in El Paso v. Argentina, in which the
tribunal favored a more ―balanced‖ view that accounts for ―both State sovereignty and the State‘s responsibility to
create an adapted and evolutionary framework for the development of economic activities, and the necessity to
protect foreign investment and its continuing flow.‖ El Paso Energy International Company v. Argentine
Republic, Decision on Jurisdiction, No. ARB/-3/15, paras. 69–70. See also UNCTAD, INTERNATIONAL
INVESTMENT RULEMAKING, para. 34, U.N. Doc. TD/B/COM/2/EM.21/2 (2007) (―Mostly as a result of the
enormous increase in investment disputes in recent years, discussion of what should be the counterweight to
investors‘ rights has gained momentum.‖).
167. SGS v. Philippines, ICSID No. ARB/02/6, para. 117.
168. Id. para. 118.
169. See supra Part IV.A.2.
170. SGS v. Philippines, ICSID No. ARB/02/6, para. 124.
171. Id.
172. Id. para. 141.
173. Id. para. 52.
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 545
interpretation does not necessarily lead to an evisceration of the dispute settlement
clauses.174
The remainder of the decision dealt with the effect of umbrella-clause-based ICSID
jurisdiction upon the forum-selection clause. The tribunal began from the principal ―that a
binding exclusive jurisdiction clause in a contract should be respected, unless overridden by
another valid provision.‖175
The relevant provision of the CISS Agreement between SGS
and the Philippines provides: ―All actions concerning disputes in connection with the
obligations of either party to this Agreement shall be filed at the Regional Trial Courts of
Makati or Manila.‖176
First, the tribunal determined whether the forum-selection clause was overridden by
the BIT provision that provided for ICSID arbitration. Concluding that the BIT did not
override the contract clause, the tribunal cited two reasons. First, the BIT clause in which
the parties consent to ICSID jurisdiction is drafted in general language, negating any
presumption that it ―has the effect of overriding specific provisions of particular contracts,
freely negotiated between the parties.‖177
Further, BITs were negotiated between countries
in order to enhance and protect investment contracts, so the BIT‘s effect cannot be to
replace in substance those same contracts.
Second, the question was raised whether Article 26 of the ICSID Convention could
override the CISS Agreement‘s forum-selection clause. SGS argued that when it began
ICSID arbitration proceedings against the Philippines in 2002, both parties thereby
consented ―to such arbitration to the exclusion of any other remedy,‖ as provided in ICSID
Article 25(1).178
The tribunal characterized SGS‘ request for arbitration as an agreement to
ICSID arbitration between two parties, SGS and the Philippines.179
Although the tribunal
acknowledged that such an agreement could in theory override the previous CISS
Agreement between the same two parties, the forum-selection clause was nevertheless
preserved.180
The tribunal supported this conclusion for three reasons. First, Article 26 ―was
intended as a rule of interpretation, not a mandatory rule.‖181
As a result it could not have
the legal effect ascribed to it by SGS. Second, Article 25 provides that consent to ICSID
arbitration ―shall, unless otherwise stated, be . . . to the exclusion of any other remedy.‖182
This indicates that the exclusivity of the remedy in Article 25(1) is subject to modification,
and in fact, SGS and the Philippines effected such a change when they agreed to a contrary
provision in the CISS Agreement, which provided for dispute resolution in the Philippine
courts.183
Third, the tribunal found SGS‘ Article 25(1) argument unattractive because it
would make the parties‘ legal rights dependent on the forum selected by the investor under
the BIT.184
If the investor were to choose an arbitration like ICSID, Article 25(1) would
override the contract. If instead another arbitral institution were selected, the parties
contractual obligations would be unaffected if that institution‘s rules had no 25(1)
174. Id. para. 123.
175. Id. para. 138.
176. SGS v. Philippines, ICSID No. ARB/02/6, para. 137.
177. Id. para. 141.
178. Id. para. 144.
179. Id. para. 16.
180. Id. para. 145.
181. Id. para. 146.
182. SGS v. Philippines, ICSID No. ARB/02/6, para. 147.
183. Id.
184. Id. para. 148.
546 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
equivalent.185
For all of these reasons, the forum-selection clause remained intact and was
given its full effect.
2. Analysis
The significance of SGS v. Philippines rests on its twin outcomes: accepting
jurisdiction over contract-based claims and deciding to give effect to the contract‘s forumselection
clause despite that jurisdiction.
A number of ICSID tribunals have adopted the Pakistan approach,186
but others side
with the Philippines tribunal by accepting jurisdiction. For example, in L.E.S.I.-DIPENTA
v. Algeria, the tribunal accepted jurisdiction over a contractual claim on account of a BIT
umbrella clause.187
Eureko B.V. v. Poland found the imperative wording and broad scope of
the BIT determinative in finding that a BIT encompassed all of a state‘s obligations,
including investor-state contracts.188
The tribunal in Sempra Energy Int’l v. Argentina
found that the claims were simultaneously contract claims and treaty claims. The umbrella
clause served to strengthen this connection, the tribunal decided, based on ―the contract, the
context of the investment and the Treaty.‖189
The Noble Ventures v. Romania tribunal
decided that an umbrella clause clearly references contracts entered into between the host
state and the investor.190
The CMS v. Argentina tribunal, discussed above, gave effect to the
BIT‘s umbrella clause despite a relatively narrow reading.191
Other ICSID tribunals have
also accepted jurisdiction over contract claims based on a BIT‘s umbrella clause.192
While the issue of jurisdiction over contract claims is still being debated, a new
problem has arisen, as predicted by the Pakistan tribunal. Indeed, the debate may be
shifting from the desirability of accepting jurisdiction over contract claims to whether BIT
provisions allowing for ICSID jurisdiction override the forum-selection clauses in investorstate
contracts.
By staying the proceedings pending resolution of the contract claims in a Philippine
court, the Philippines tribunal attempted to allay Pakistan‘s concerns of an umbrella clause
overriding contract forum-selection clauses.193
Despite accepting jurisdiction over the
contract claims, the tribunal did not decide the merits. The dissenting arbitrator in
Philippines disagreed with this part of the result, advocating a decision on the merits based
on the broad reach of the umbrella clause and consent to ICSID arbitration in the BIT.194
Many advocate a position similar to the Philippines dissent that would always give a
BIT‘s dispute resolution provision precedence over the investor-state contract. If this
185. Id.
186. See supra Part IV.A.2.
187. Consorzio Groupement L.E.S.I.-DIPENTA v. People‘s Democratic Republic of Algeria, Award, ICSID
Case No. ARB/03/8 (2005).
188. Eureko B.V. v Republic of Poland, Partial Award on Liability paras. 255, 257 (ad hoc arbitration of
Aug. 19, 2005), available at http://ita.law.uvic.ca/documents/Eureko-PartialAwardand DissentingOpinion.pdf.
189. Sempra Energy Int‘l v. Argentine Republic, Decision of the Tribunal on Objections to Jurisdiction,
ICSID Case No. ARB/02/16, para. 101 (2005).
190. Noble Ventures, Inc. v. Romania, Award, ICSID Case No. ARB/01/11, paras. 59–61 (2005), available
at http://ita.law.uvic.ca/documents/Noble.pdf.
191. See supra note 139 and accompanying text.
192. See LG&E Energy Corp. v. Argentine Republic, Decision of the Tribunal on Objections to Jurisdiction,
ICSID Case No. ARB/02/16, para. 46 (2004); ADC Affiliate Ltd. v. Republic of Hungary, Award, ICSID Case No.
ARB/03/16, para. 294 (2006).
193. SGS v. Philippines, ICSID No. ARB/02/6 para. 177(c).
194. SGS v. Philippines, ICSID No. ARB/02/6 (dissent).
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 547
position were adopted, a country‘s consent to ICSID arbitration as an arbitration forum
would render the contract‘s forum-selection clause ineffective. This would put the entire
spectrum of investor-state disputes in ICSID‘s domain.
But as SGS v. Philippines indicates, just because jurisdiction is accepted does not
mean the dispute settlement clauses in individual contracts are going to be nullified.195
To
the extent the Pakistan tribunal, or any tribunal, is worried about eviscerating a contract
forum-selection clause, it need not exercise jurisdiction once accepted.
Is there a sound basis for the Philippines approach, when it is customary practice in
arbitration tribunals to decide the merits after accepting jurisdiction? The tribunal‘s
decision to accept jurisdiction but pass on the merits of the contract claim deserves special
scrutiny, because it defies expectations of how the arbitral process usually works. It is
worthwhile to determine the validity of the tribunal‘s reasons for such a holding and inquire
as to whether alternative reasons may exist for accepting the SGS v. Philippines result.
The tribunal advances two primary arguments to support its position. First, even
though the contract claim is elevated to the level of a BIT claim, it cannot hide its origins.
In most cases, the fact that the claim is contractual compels resolution of the claim in the
forum chosen by the contract. At the very least, the tribunal ―cannot accept that standard
BIT jurisdiction clauses automatically override the binding selection of a forum by the
parties to determine their contractual claims.‖196
The tribunal takes a significant step: it
employs the Vivendi fundamental basis to support the conclusion that contract claims should
be adjudicated in accordance with the contract‘s forum-selection clause. It will be
remembered that while the Vivendi annulment committee required the tribunal in that case to
decide a claim whose fundamental basis is in the BIT, it was left open to tribunals to decide
what to do with a fundamentally contractual claim. ―[W]here the essential basis of a claim
brought before an international tribunal is a breach of contract,‖ the committee stated, ―the
tribunal will give effect to any valid choice of forum clause in the contract.‖197
When it
comes to deciding the merits of SGS‘ contract claim, the tribunal cannot ignore the contract
because it provides the claim‘s fundamental basis.
The valid use of Vivendi by the tribunal in Philippines contrasts sharply with the
fundamental basis test as employed in Pakistan. In that decision, the tribunal, before even
reaching the jurisdictional issue created by the umbrella clause, decided whether it had
jurisdiction over SGS‘ BIT claims for expropriation and unfair treatment. The tribunal cited
Vivendi and the fundamental basis test to support accepting jurisdiction over the BIT claim.
As previously noted, this is problematic for two important reasons. First, the Vivendi test
was never meant to be employed before a tribunal decides to accept jurisdiction. Instead, as
in both Vivendi and Philippines, the proper role of the fundamental basis test is to guide the
tribunal in its decision on the merits of a BIT-based claim. If not BIT-based, the tribunal
can give effect to the contract‘s choice of forum. Second, as a result of its misapplication of
Vivendi, the Pakistan tribunal seemed to have already made up its mind about contractbased
claims by the time it examined the BIT‘s umbrella clause. By limiting BIT claims to
those with a fundamental basis was in the BIT, the tribunal closed the door on any argument
that would elevate the contract claim to the same level as a BIT claim. This is made all the
more puzzling by the tribunal‘s decision to not inquire too closely at the jurisdictional stage
of the proceedings as to whether a claim was based in contract or treaty.
195. Id. para. 123.
196. Id. para. 153.
197. Id.; Vivendi v. Argentina, ICSID No. ARB/97/3, para. 98 (2002).
548 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
If the SGS cases were to be judged merely on the basis of which tribunal interpreted
Vivendi correctly, there would be less discussion about the irreconcilable conflict between
the SGS cases, and a wider consensus as to the correctness of the Philippines holding. One
of this article‘s purposes has been to dismantle the perceived conflict between the SGS cases
on the issue of jurisdiction over contract-based claims. The desired result is a renewed
focus on the controversy surrounding ICSID tribunals once jurisdiction is accepted.
The Vivendi test provides a healthy dose of support for the Philippines result. The
Philippines tribunal knew, however, that implementing the Vivendi standard, without
something more, would not satisfy critics. A previous ICSID decision, while useful for
reference and to support a tribunal‘s argument, is not binding precedent.198
Therefore, the
tribunal searches for an argument to supplement the Vivendi approach. No doubt it realized
the novelty of accepting jurisdiction over a claim but then letting another forum decide it.
The solution settled on by the tribunal distinguishes between ―jurisdiction‖ and
―admissibility.‖
In its second main argument, the tribunal says that it did not accept jurisdiction over
SGS‘ contract claims; rather, the claims were admitted, at which time the fundamental basis
of the claims required resolution elsewhere. Finding a claim to be admissible, but then
leaving resolution to another forum, is the same approach embodied, for example, in the
forum non conveniens doctrine.199
Whether or not such a maneuver is justified, using it as a
basis for the Philippines decision lends little clarity to the debate over whether or not
contract claims should be decided by an ICSID tribunal. The tribunal relies on the principle
that ―a party to a contract cannot claim on that contract without itself complying with it‖ and
decides that such a restriction ―is more naturally considered as a matter of admissibility than
jurisdiction.‖200
If an investor wants to make a contract-based claim against the host
country, it must do so consistent with its obligations under the contract.201
One response to the SGS v. Philippines decision would demand that the tribunal, if it
will not review the contract claim once admitted, refuse the contract claim at the outset.
Another view is that favored by the Philippines dissent: once a contract claim is admitted,
the merits should be decided as a matter of course. These opposite approaches are neither
necessary nor in accordance with principles underlying international investment law. The
Philippines outcome is desirable, because it relieves ICSID tribunals from deciding each
and every contract claim, while ensuring that ICSID tribunals have the discretion to decide a
case if necessary. The challenge is to defend the Philippines result without relying on the
admissibility classification that is the basis for the decision. The next section of the article
offers suggestions that would lead to the Philippines result based on principles of
international investment law.
198. SGS v. Philippines, ICSID No. ARB/02/6 para. 97 (acknowledging that while consistency with other
ICSID tribunals is preferable, decisions may be inconsistent due to different applicable law and the lack of
precedent in international law).
199. See Piper Aircraft v. Reyno, 454 U.S. 235 (1981).
200. SGS v. Philippines, ICSID No. ARB/02/6, para. 154.
201. Another route to the same result would find the investor-state contract to be a waiver by the investor of
its right under the BIT to bring a contract claim before ICSID. Assuming that the contract claim is elevated to a
BIT claim by the umbrella clause, the investor may have waived his right to an ICSID forum upon becoming a
party to the contract with the host state. ICSID arbitration is not absolute even for substantive BIT claims. See
supra note 100 (discussing the common BIT ―fork in the road‖ provision, precluding an investor from choosing
ICSID arbitration after submitting a BIT claim to the courts of another contracting party). If triggered, the ―fork in
the road‖ would close the door to ICSID arbitration for substantive BIT claims and contract claims alike. The
arguably more attenuated contractual BIT claim would be considered waived as soon as the investor enters a
contract containing an alternate forum-selection clause. By becoming a party to the contract, the investor would
be limited to the dispute resolution provision of the contract, to the exclusion of ICSID arbitration.
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 549
V. EVALUATING THE ROLE OF ICSID TRIBUNALS: THE SHORTCOMINGS OF
A ―DECIDE EVERYTHING‖ APPROACH
A. Consent Controls: The BIT as an Agreement to Arbitrate
The Switzerland-Philippines BIT is just one of the thousands of BITs in which
contracting parties consent to ICSID jurisdiction. What is more, the choice of whether to
arbitrate before ICSID or some other forum is ultimately in the hands of the investor who
submits the request for arbitration. Although ICSID is the forum favored by investors, the
possibility remains that an alternate forum, such as an UNCITRAL or ICC tribunal, will
decide the dispute. Upon signing a BIT, a host country has not only consented to arbitrate;
it has also consented to be bound by the investor‘s choice of forum. Therefore, when the
investor submits a request for arbitration, the host country is getting just what it bargained
for in the BIT. There is no element of surprise when an investor brings a claim, since the
country agreed to arbitrate with investors and no one else. Furthermore, the host country
cannot claim altered expectations when a given arbitral forum is chosen, because it
relinquished that choice to the investor in the BIT.202
In short, the consent given by the host
country is the widest kind of consent possible. Packaged within this consent are three
important consequences, each following from the other.
First, the host country has consented to have arbitrators determine questions of
jurisdiction. In other words, there is an acknowledgment in the BIT that only the arbitrators
are in a position to decide the claims over which they have jurisdiction. Thus, the threshold
question of who decides the jurisdiction of the arbitrators is for the arbitrators themselves.
This is in part a result of the nature of international arbitration, in which arbitrators are the
only actors in a position to make such a decision.
This conclusion is only enhanced by the dispute resolution provisions in BIT. By
consenting in the BIT to resolve claims before ICSID, a host country has necessarily vested
in the tribunal the power to decide its own jurisdiction. As a result, the opposing sides in the
SGS v. Philippines decision could be described as doing nothing more than trying to
persuade the tribunal that it did or did not have jurisdiction. Accordingly, it is difficult to
conclude that the tribunal did not reach the correct result, at least in the sense that it was the
tribunal‘s decision to decide whether it had jurisdiction over SGS‘ contract claims. In
ICSID decisions, the issue of who decides questions of jurisdiction is rarely contested.
What has not been emphasized enough in the current commentary on the SGS cases is the
enormous effect the BIT parties‘ consent has on questions of jurisdiction. In the BIT, both
countries have have consented to have the question of ―who decides‖ resolved by whatever
arbitrator is chosen by the investor.
Second, once it is acknowledged that ICSID arbitrators determine their own
jurisdiction, it becomes even easier to accept the holding in SGS v. Philippines that the
umbrella clause elevates a contract claim to the level of a BIT claim. The reasons for a
broad reading of the umbrella clause have already been explored in some detail. A further
basis can be found in Lanco v. Argentina. In that decision, the tribunal found that it had
jurisdiction over an investor‘s BIT claims notwithstanding the forum-selection clause in the
202. See SGS v. Philippines, ICSID No. ARB/02/6, para. 50 (citing SGS‘ argument that ―[t]he Philippines
have consented to ICSID arbitration as one of the available means of dispute resolution and the final decision as to
whether . . . ICSID arbitration supersedes [the contract‘s forum-selection clause] rests solely in the hands of the
investor.‖).
550 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
investor-state contract. More important than the holding itself is the rationale given by the
tribunal. It was the consent given by Argentina in the BIT that persuaded the tribunal to
hold fast to the jurisdictional provision in the BIT, despite the contract‘s forum-selection
clause.
The SGS v. Philippines decision would have been bolstered by a discussion
emphasizing the consent the Philippines gave in the BIT to arbitrate disputes before ICSID.
The tribunal‘s argument in Lanco was that consent to arbitrate BIT claims overrode all else.
This reasoning could have been transposed to the SGS v. Philippines facts. The Philippines,
having consented to arbitrate a wide array of disputes before ICSID, would have to
overcome a presumption that they consented to arbitrate contract claims as well, especially
given the presence of an umbrella clause. For just as Argentina‘s consent in Lanco was
undiminished by the contract, the Philippines‘ consent should be accorded a great deal of
weight in determining whether it agreed to arbitrate contract claims. In Lanco, the tribunal
reasoned that ―when the parties give their consent to ICSID arbitration, they lose their right
to seek to settle the dispute in any other forum, domestic or international . . . .‖203
The fact
that the Philippines at least agreed to arbitrate something with a Swiss investor before ICSID
should be given due consideration in determining whether an umbrella clause is broad
enough to include contract claims within that consent. To be sure, this would not have
conclusively demonstrated ICSID‘s jurisdiction over the contract claims, but explaining the
result in terms of the host country‘s consent given in the BIT would have added another
level of legitimacy to the Philippines holding.
Finally, if a tribunal is able to decide whether it has jurisdiction over a claim, and if it
decides that jurisdiction exists, the next question is the extent to which an arbitrator is
empowered to adjudicate a claim. This is in some ways the most straightforward power
vested in an arbitration tribunal. The parties‘ consent in the BIT must be read like any other
agreement to arbitrate. As a result, both the investor and the host country have by mutual
consent agreed in advance to privatize the resolution of their dispute. Although it is unclear
whether BIT parties could subject an ICSID award to review by a national court, this is not
an issue at present, since no BITs provide for such judicial review. What is clear enough,
however, is that consent given in the BITs extends to the final decision of the tribunal.
The decision in SGS v. Philippines to stay the proceedings pending the outcome of a
contract claim in Philippine courts has been attacked as an abdication of arbitral
responsibility.204
But from the standpoint of consent, it is difficult to see why this is so. The
arbitrators, entrusted to resolve the dispute before them, are vested with the power to render
any judgment they deem appropriate, within the limits of the ICSID Convention.205
The
decision to allow resolution of contract claims by the courts of the host country, while
perhaps not the outcome expected by the parties, is nevertheless an outcome to which they
consented. Viewing the BIT‘s forum-selection clause as an agreement to arbitrate suggests
that basic principles of arbitration should control. Chief among those principles is party
autonomy. It has been said that arbitration, as a creature of party autonomy, is ―at bottom
little more than ‗the parties‘ dream.‘‖206
Like any arbitration, therefore, an ICSID
arbitration is subject to change by the parties when they agree to arbitrate. Subject as it is to
alteration, the agreement could stipulate that a tribunal shall decide, contrary to Philippines,
203. Lanco v. Argentina, ICSID Case No. ARB/97/6, para. 36.
204. Wong, supra note 16, at 137–38.
205. See ICSID Convention, supra note 3, art. 52 (providing very limited grounds for annulment).
206. Alan Scott Rau, The Culture of American Arbitration and the Lessons of ADR, 40 TEX. INT‘L L. J. 449,
449 (2005) (quoting HENRY M. HART, JR. & ALBERT M. SACKS, THE LEGAL PROCESS: BASIC PROBLEMS IN THE
MAKING AND APPLICATION OF LAW 336 (tent. ed. 1958)).
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 551
every claim over which it accepts jurisdiction. Absent such alteration, however, the tribunal
should not be held to such a restriction.
B. Certainty in an Uncertain World? Choices Facing ICSID Tribunals
The effect of the SGS v. Philippines decision was to leave resolution of the claimant‘s
contract claims to the forum designated by the contract. This result has given rise to various
grounds for criticism, the most important of which—that uncertain decisions fail to
safeguard investors‘ expectations—will be discussed last.
The first criticism questions the Philippines tribunal‘s insistence on separating the
contract claims from treaty claims. The argument is that both the BIT and the investor-state
contract have broadly worded dispute resolution clauses. For example, the BIT provides for
resolution of all disputes before an international tribunal, and the contract provides for all
disputes to be resolved in the courts of the host country. Neither instrument allows for the
possibility of settling contract disputes in one forum and treaty disputes in another.
Therefore, the Philippines tribunal had no basis for treating the claims separately: ―Nothing
in these provisions contemplates, much less authorizes, the determination of different
components of the dispute (whatever they may be) in different fora.‖207
Such an argument at least helps frame the debate by highlighting the tension between
the BIT and the contract. Both are valid agreements, negotiated at arms‘ length. In the BIT,
the host country agrees to arbitrate a wide range of disputes in a forum chosen by the
investor. In the contract, the investor agrees to resolve all contractual disputes in the
designated forum. Therefore, two valid agreements demand recognition by an ICSID
tribunal. What should be clear enough is that an ICSID decision on the merits of a contract
claim runs afoul of the parties‘ agreement in the contract. There are defensible positions for
interpreting ICSID‘s jurisdiction broadly enough to permit this.
But an argument that disallows an ICSID tribunal from distinguishing between contract
and treaty claims, simply because neither instrument contemplates such a distinction, does
little to suggest on what basis a tribunal may decide any claims. For by the same reasoning,
a tribunal should refuse to decide both BIT and contract claims because ―determination of
different components of the dispute‖ is unauthorized. A better approach focuses on the
validity of both agreements, underscoring the dilemma facing the Philippines tribunal: that
a recognition of either agreement invariably violates the other. The tribunal found
persuasive the Philippines‘ argument that SGS ―may not seek to enforce the ‗specific‘
payment obligations of the CISS Agreement whilst at the same time disregarding the
agreement to resolve these payments (and all other contractual) obligations in [the
Philippine courts].‖208
Determining the circumstances in which one of the agreements may
be violated requires an analysis of the policies underlying international investment law. An
inevitable result of such an inquiry is what critics of SGS v. Philippines have termed
―inconsistent‖ results.
Inconsistency is a second perceived weakness in the Philippines holding, which
according to one commentator ―results in the BIT tribunal having jurisdiction over an empty
shell and deprive[es] the BIT dispute resolution provision of any meaning.‖209
By accepting
207. Wong, supra note 16, at 168.
208. SGS v. Philippines, ICSID No. ARB/02/6 para. 59.
209. Emmanuel Gaillard, Investment Treaty Arbitration and Jurisdiction over Contract Claims: The SGS
Cases Considered, in INTERNATIONAL INVESTMENT LAW AND ARBITRATION 334, 344–45 (Todd Weiler ed.,
2005).
552 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
jurisdiction but staying the proceedings, the tribunal renders the umbrella clause
―superfluous.‖210
It guarantees a decision regarding the contract claims only in ―extracontractual‖
circumstances, such as fraud, a ―mis-carriage of justice,‖ or circumstances
rendering contract performance impossible. For these reasons, the Philippines stance is
described as internally ―inconsistent.‖ The proposed solution? If an umbrella clause is to
have any effect, its scope must be absolute, resulting in complete ICSID jurisdiction over
contract claims. Anything less causes uncertainty for the investor, who will not know what
to expect from such ―inconsistent‖ results. According to those favoring ICSID arbitration of
contract claims, uncertainty in international arbitration must be avoided at all costs.
To say that accepting jurisdiction over contract claims without deciding them is
―inconsistent‖ is self-evident, but only if one accepts at the outset that consistency requires
deciding all claims over which jurisdiction is accepted. Determining, in turn, what amounts
to consistency requires an examination of the applicable arbitration rules and, next, any
accepted arbitration norms and practices.211
Since the ICSID arbitration rules are silent on
whether a tribunal must decide the merits of a claim once accepted, other sources must be
consulted.
It is tempting to find a solution to the problem in the Vivendi annulment decision. In
that case, the committee determined that the tribunal had manifestly exceeded its powers by
refusing to decide BIT claims whose jurisdiction had already been determined. Using
similar reasoning, one might argue that a refusal to decide contract claims whose jurisdiction
has already been determined is likewise unacceptable. But this would ignore the nature of
the claim in Vivendi, which was based on the BIT‘s substantive provisions. The tribunal had
no reason to refuse to decide the merits of the BIT claims, because the BIT itself provided
for ICSID jurisdiction. In contrast, by refusing to decide contract claims, the Philippines
tribunal was merely giving effect to the contract‘s dispute resolution clause. Therefore,
there is no easily identifiable basis for labeling the Philippines decision inconsistent,
because such a label raises the common-sense question: ―Inconsistent with regard to what?‖
The real basis of the inconsistency argument seems to be the discomfort many have
with the novel result reached in SGS v. Philippines. Based on voluminous systems of
national law and years of precedent, we have come to expect that once a court accepts
jurisdiction over a claim, it will decide the merits. Only in certain circumstances will a court
accept jurisdiction but have the merits decided elsewhere (for example, by employing the
doctrine of forum non conveniens). By extension, this widespread expectation has been
applied to arbitration tribunals without engendering much controversy. ICSID tribunals
bear a likeness to arbitration tribunals generally. For example, they are creatures of
agreement governed by principles of party autonomy, and they represent the fruits of the
parties‘ decision to privatize the resolution of disputes.
But for all of the similarities, ICSID arbitration is unique. More than the result of a
bargain struck between private parties, it is a product of the multilateral ICSID Convention
and a bilateral treaty whose purpose is to protect investors.212
Furthermore, ICSID obtains
its jurisdiction from the consent of the same countries who sign BITs for the sole purpose of
attracting the capital of private investors. As investments in developing countries continue
210. Wong, supra note 16, at 167.
211. See Tai-Heng Cheng, Power, Authority and International Investment Law, 20 AM. U. INT‘L L. REV.
465, 470 (2004) (―Arbitral tribunals and international courts have usually been custodians of international
investment norms when called upon by investors and states to adjudicate investment disputes.‖).
212. See Franck, supra note 6, at 1538 (―[Shifts made by investment treaties] created a mechanism to bolster
investors‘ confidence that they will receive a ‗fair shake‘ when resolving disputes with Sovereigns, thus reducing
the risks associated with investment and, arguably, increasing the incentive to invest abroad.‖).
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 553
to increase, and as countries naturally become more involved with these investments,
investor-state contracts will become increasingly prevalent and will have to coexist
alongside BITs.
Only ICSID is in the position of balancing these kinds of mixed public-private
agreements involving contradictory dispute resolution provisions. It should be noted that, in
addition to the competing provisions found in BITs and contracts, ICSID tribunals are faced
with different laws governing the agreements. This is because the BIT is a product of
international law and the contract is governed by the law agreed upon by the investor and
the host country. Therefore, as a basis for dismissing the Philippines result, an argument
based on inconsistency fails to appreciate the divergent interests implicated when an
investor brings a contract claim against the country of investment and the difficult balancing
act such a claim requires of ICSID tribunals.
Finally, there is a policy argument that seeks uniformity of result in order to protect
investors‘ expectations. This seems to be the most commonly accepted argument for giving
ICSID broad jurisdiction that would include contract claims. The primary impetus behind
the ICSID Convention was the protection of investors, who are powerless to compel a
sovereign to settle a dispute absent provisions in a treaty or convention.
The SGS tribunal got it halfway right by reading the umbrella clause broadly, but the
international investment policy demands that a tribunal decide the merits of all investment
claims. Otherwise, the uncertain prospects facing the investor will discourage investment.
In short, this line of thinking says that, if there is any kind of protection for an investor, it
must be absolute.
The protection of investor expectations is a worthwhile goal, and to the extent ICSID
tribunals weigh policy concerns, investor protection should be accorded its due weight.
Uncertainty has the potential to slow investment, and it is ICSID‘s purpose to safeguard
investors expectations. However, it is not clear that the resolution of contract-based claims
by ICSID will significantly alter investors‘ expectations.
There is little evidence that by increasing its role to include contract claims, ICSID
would provide more certainty to the investor. Current commentary often takes for granted
the fact that deciding contract claims will affect expectations.213
One practitioner concludes
that ―more aggressive protections, such as elevating all investment undertakings into treaty
obligations, are often not critical to promoting investments because an investor‘s basic
investment is secure even without such aggressive protections.‖214
In other words, an
investor‘s behavior may be little-affected by whether or not ICSID decides contract claims.
In large part, this is due to the relatively limited importance of contract claims to an
investor‘s overall decision-making process. Of primary concern to the investor is whether
the host country is party to a BIT under which the investor has a claim. Increased
protections ―should be viewed with some suspicion because any positive effect these
protections may have on global investments is insufficiently supported by evidence.‖215
At
present, the international investment environment affords no basis to conclude that ICSID
must decide contract claims to spur investment.
Even if it could be shown that certainty is enhanced by such measures, other
mechanisms are in place to safeguard investor expectations. The most important safety net
for investors is still the BIT, which provides basic yet indispensable protections against
213. See, e.g., Franck, supra note 6.
214. Tai-Heng Cheng, supra note 211, at 502–03.
215. Id. at 504.
554 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
expropriation and unfair treatment.216
When an investor enters into a contract with the host
country, the BIT claim remains at the investors‘ disposal. By itself, such a treaty transforms
investors into ―private attorneys general‖ able to enforce a BIT‘s international law standards
against a sovereign.217
Nor is international arbitration the only mechanism for reducing an
investors‘ risk. Subrogation of an investor‘s claim by its home country is another widely
used method.218
In addition, protection against detrimental action by the host country may be
provided in the contract itself, in the form of a stabilization clause, an exchange control
clause, or so-called denationalization of development agreements.219
If the resolution of
contract claims by ICSID has an effect on an investor‘s calculations, this effect is
diminished by other mechanisms in place to safeguard investor expectations.
Perhaps most significantly for the present discussion, the SGS v. Philippines result
leaves open the possibility of protecting an investor‘s contract-based claims under certain
circumstances.220
Even the decisions rejecting the Philippines result allow that examination
of an investor‘s contract claims will be warranted in the event that a contract breach is so
egregious as to rise to a breach of an international law standard.221
The fact that an
investor‘s contract claim will at least pass through the hands of the tribunal should be of
some comfort to an investor calculating risks.
The trend in decisions after Philippines points to a likelihood that an ICSID tribunal
will give itself the option to decide the merits of a contract claim.222
This reflects a positive
consequence of the Philippines decision: tribunals that disagree with the reasoning can
nevertheless decide the merits of a contract claim, while maintaining fidelity to the result in
the Philippines decision simply by accepting jurisdiction. A tribunal need only declare that
it has found a reason to override the contract that was lacking in Philippines. Leaving it to
individual tribunals to decide when an investor-state contract‘s forum-selection clause
should be violated at the expense of the BIT is preferable to mandating the resolution of
contract claims in every circumstance. Such an ad hoc approach is not only in line with the
tribunal‘s purpose, it reflects an understanding that deciding contract claims may or may not
be necessary to safeguard investor expectations, given the particulars of each case. In view
of the multitude of risk-minimizing measures at the investor‘s disposal, it is safe to conclude
that in many cases, resolution of contract claims has minimal effect on investor expectations.
Finally, supposing it could be demonstrated that certainty for investors would be
markedly enhanced, I submit that investor certainty is an insufficient reason for ICSID to
arbitrate contract claims. While desirable, certainty should not be allowed to obscure the
difficult issue facing ICSID tribunals in the form of two valid agreements, the BIT and the
investor-state contract, each with its own forum-selection clause. This type of conflict is the
inevitable result of increasing international investment, and it will only become more
common. The constantly evolving nature of international investment has already proved too
complex to be resolved by treaty, the instrument of choice of sovereigns. The result is a
relinquishment of power from public to private actors. Discussing this phenomenon, TaiHeng
Cheng concludes: ―The power and authority that international investment law drains
from states does not vaporize, and is often transferred to a wide range of decision-makers
. . . . Among these transferees, the greatest beneficiaries are foreign and international
216. Id. at 502.
217. Franck, supra note 6, at 1538.
218. See DOLZER & STEVENS, supra note 1, at 156–64.
219. See Piero Bernardini, Development Agreements with Host Governments, in ECONOMIC DEVELOPMENT,
FOREIGN INVESTMENT AND THE LAW 161, 170–72 (Robert Pritchard ed., 1996).
220. SGS v. Philippines, ICSID No. ARB/02/6 para. 154.
221. See Joy Mining v. Egypt, Award on Jurisdiction, ICSID Case No. ARB/03/11 para. 75 (2004).
222. See supra notes 186–166 and accompanying text.
2008] SGS V. PHILIPPINES AND THE ROLE OF ICSID TRIBUNALS 555
tribunals and investors.‖223
The more complex the issue before an international institution
such as ICSID, the less useful it is to question the outcomes it reaches, due to the lack of
alternative solution mechanisms. ICSID is a vanguard institution in the new global economy
in which states increasingly cede decision-making power. ―Many of the new issues,‖ writes
one scholar, ―are not only ‗global‘ but they are also somehow irresistible, in the sense that
they are outside the control of national authorities.‖224
Advocating that an ICSID tribunal should always decide contract claims to ensure
certainty misses the mark, and not just because certainty cannot be guaranteed. Global
integration has further convinced investors that the rewards of economic development
outweigh the comfort of certainty. Though investors will gladly accept any new protections
granted by ICSID or by individual BITs, they understand that international investment will
always pose risks. Among these risks at present is the possibility that contract claims will be
resolved in the host country, even when the claim is elevated to the same level as a BIT‘s
substantive protections. It is unclear the extent to which this realization affects the behavior
of investors, if at all. But an approach requiring the resolution of contract claims by ICSID
tribunals in all cases does little to untangle the competing forum-selection provisions found
in BITs and investor-state contracts. Different forum-selection clauses governing the same
party relationship is an inevitable result of international investment if there ever was one.
VI. CONCLUSION
The SGS v. Philippines decision, while viewed favorably by many compared to the
narrow umbrella clause reading in SGS v. Pakistan, has been criticized for not deciding the
merits in the underlying dispute between the parties. This article has defended the result in
SGS v. Philippines on two grounds. First, the consent of the parties, evidenced in the BIT,
transfers a substantial amount of decision-making power to the ICSID tribunal. Requiring
that a tribunal decide the merits of every claim over which it has jurisdiction ignores this
consent. Second, the Philippines middle ground is the only way to respect both the integrity
of the bilateral investment treaty and the contract between investor and host state. While
such a result is not perfect, ICSID tribunals should not alter their approach in order to
safeguard investors‘ expectations. ICSID jurisdiction over an investor‘s BIT claims and
contract claims is all the certainty an investor can and should be able to expect in an
inherently risky enterprise.
As international investment evolves and dispute resolution matures, private decision
makers like ICSID tribunals will be given more autonomy. The inability of local courts to
resolve certain international disputes, recognized by BITs and subsequently by ICSID
arbitration, means that a leaner solution to problems arising in international investment
relations is more necessary than ever. In answering ―whether human law should be changed
in any way,‖ St. Thomas Aquinas gave an answer that resonates today:
It seems natural to human reason to advance gradually from the imperfect to the
perfect. . . . So also in practical matters: for those who first endeavored to
discover something useful for the human community, not being able by
themselves to take everything into consideration, set up certain institutions which
223. Tai-Heng Cheng, supra note 211, at 492.
224. Ambassador Emilio J. Cárdenas, The Notion of Sovereignty Confronts a New Era, in ECONOMIC
DEVELOPMENT, FOREIGN INVESTMENT AND THE LAW 13, 15 (Robert Pritchard ed., 1996).
556 TEXAS INTERNATIONAL LAW JOURNAL [VOL. 43:523
were deficient in many ways; and these were changed by subsequent lawgivers
who made institutions that might prove less frequently deficient . . . .225
Arbitration is by nature an improvement over legal systems unsuited to the demands of
a global economy. Therefore, calls for ICSID tribunals to fashion decisions to ensure a
given result run counter not only to the consent of the parties, but to the very purpose of
arbitration. The open-ended and at times conflicting decisions by tribunals do not afford the
consistency that many investors seek, yet the primary purpose of international arbitration—
to impartially resolve disputes arising from transnational transactions—is superbly served by
agreements such as BITs and the arbitration clauses contained in them. As an examination
of the SGS cases illustrates, requiring a guarantee from ICSID tribunals that they will decide
the merits of contract claims exacts too high a cost: by imposing our notions of what we
think an arbitral decision should look like, we risk losing sight of the advantages the
arbitration process provides. ICSID arbitrators, because of their unique position in the everchanging
international investment landscape, should have the freedom to decide the methods
employed in the resolution of disputes over which they have jurisdiction. Bilateral
investment treaties vest in them nothing less.
225. THOMAS AQUINAS, SUMMA THEOLOGICA: FIRST COMPLETE AMERICAN EDITION 1022 (Fathers of the
English Dominican Province trans., Benziger Bros. 1947).