ArticlePDF Available

Can International Investor-State Disputes be Prevented? Empirical Evidence from Settlements in ICSID Arbitration

Authors:
ICSID Review, Vol. 29, No. 1 (2014), pp. 41–65
doi:10.1093/icsidreview/sit034 Published Advance Access December 28, 2013
Can International Investor–State Disputes
be Prevented? Empirical Evidence from
Settlements in ICSID Arbitration
Roberto Echandi
1
and Priyanka Kher
2
I. INTRODUCTION
The significant increase in international investor–State litigation activity over the past
two decades
3
has revealed a puzzling aspect of the current international investment
regime. While it is widely recognized that a mutually agreed resolution to a dispute
is more advantageous than relying on adjudication
4
—which entails the imposition of
a solution by a neutral third party—in practice, recourse to mediation or conciliation
has been quite limited in international investment disputes.
5
Literature has
attempted to identify the factors behind that trend.
6
At the same time, practice
1
Global Product Leader, Investment Policy, Investment Climate Department, World Bank Group. Email:
REchandi@ifc.org
2
Consultant, Investment Climate Department, World Bank Group. The views expressed in this study are those of
the authors and do not represent the views of the World Bank Group. The authors thank Frauke Nitschke and Eloise
Obadia for providing valuable clarifications during the research.
3
According to the United Nations Conference on Trade and Development (UNCTAD), during the past two
decades, there have been more than 500 known investor–State disputes submitted to international arbitration in
various fora (UNCTAD, ‘IIAs Issues Note, No 1 May 2013’ <http://unctad.org/en/PublicationsLibrary/web-
diaepcb2013d3_en.pdf>accessed 1 June 2013). For a comprehensive overview of the history of ICSID, see Antonio
R Parra, History of ICSID (OUP 2012).
4
Jack Coe Jr, ‘Towards a complimentar y use of conciliation in investor–state disputes: a preliminary sketch’
(2006) 21(4) Mealey’s Arb Rep 72; Jack Coe Jr, ‘Should Mediation of Investment Disputes Be Encouraged, and, If
So, by Whom and How?’ in Arthur W Rovine (ed), Contemporary Issues in Inter national Arbitration and Mediation: The
Fordham Papers (Martinus Nijhoff 2009) 339; Jack Coe Jr, ‘Settlement of Investor–State Disputes through
Mediation—Preliminary Remarks on Processes, Problems and Prospects ’ in R Doak Bishop (ed), Enforcement of
Arbitral Awards Against Sovereigns (Juris 2009) 73; Jeswald W Salacuse, ‘Is there a Better Way? Alternative Methods of
Treaty-Based, Investor–State Dispute Resolution’ (2007) 31 Fordham Intl L J 138.
5
Since its inception and until December 2012, there have been only nine conciliation cases submitted to ICSID.
Seven were registered under the Convention on the Settlement of Investment Disputes between States and Nationals
of Other States (opened for signature 18 March 1965, entered into force 14 October 1966) (‘ICSID Convention’).
Two were registered with the ICSID Additional Facility. See ICSID, ‘Caseload—Statistics 2013-1’ <https://icsid
.worldbank.org/ICSID/FrontServlet#>accessed 1 June 2013.
6
Ucheora O Onwuamaegbu, ‘The Role of ADR in Investor–State Dispute Settlement: The ICSID Experience’
(2005) 22(2) News from ICSID 12; Barton Legum, ‘The Difficulties of Conciliation in Investment Treaty Cases: A
Comment on Professor Jack C. Coe’s ‘‘Toward a Complementary Use of Conciliation in Investor–State Disputes—A
Preliminary Sketch’’’ (2006) 21 Mealey’s Intl Arb Rep 1; Mark A Clodfelter, ‘Why Aren’t More Investor–State Treaty
Disputes Settled Amicably?’ in UNCTAD, Investor–State Disputes: Prevention and Alternatives to Arbitration II,
Proceedings of the Washington and Lee University and UNCTAD Joint Symposium on International Investment
and Alternative Dispute Resolution, UNCTAD/WEB/DIAE/IA/2010/8 (29 March 2010) 38; Thomas W Wa¨lde,
‘Pro-active Mediation of International Business and Investment Disputes Involving Long-Term Contracts:
From Zero-Sum Litigation to Efficient Dispute Management’ (2004) Transnatl Dispute Management 38
ßThe Author 2013. Published by Oxford University Press on behalf of ICSID. All rights reserved.
For permissions, please email: journals.permissions@oup.com
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
clearly shows that negotiation as a means to resolve investor–State disputes is already
happening. Paradoxically, negotiations are not taking place in an amicable setting,
but rather framed within the litigious context of international investor–State
arbitration.
The main objective of this study is to get a better understanding of the profile of
settlements in arbitration disputes registered at ICSID. By analysing a series of
variables, we purport to determine on the basis of empirical evidence, whether and
to what extent ICSID arbitration is dealing with disputes that should not have
escalated to the international level in the first place, because they reasonably could
have been mutually resolved by the parties at an earlier stage.
7
To address this question, this research will examine some key contextual
variables of ICSID arbitration, including how early or late in the arbitration
proceedings settlements have tended to take place. Our hypothesis is that if
settlements are taking place so early that the parties have not used the arbitration
process to assess the strength of their respective claims and arguments, then the
need for a tribunal to elucidate a disagreement of law or fact would be
questionable. In those situations, we argue that it would be the very availability of
arbitration that is the key factor promoting settlements. Thus, the number of
disputes settled at very early stages of the arbitration process—that is, before or by
the time the tribunal is constituted—would be a good approximation of the
minimum number of disputes that could have been mutually resolved by the parties
before they escalated to international arbitration in the first place.
In particular, this study examines arbitration proceedings that were discontinued
after being registered with the ICSID Secretariat under the ICSID Convention
8
or
under the ICSID Additional Facility Rules
9
between 1970 and 2012.
10
The
research focused on the frequency of settlements,
11
the legal ground invoked by
the parties to discontinue the arbitration proceedings, at what point during the
arbitration process the settlements took place, the settlements by economic sector
and a brief characterization of the countries that were involved in settlements
during the period under study.
From the outset it should be clarified that this study does not attempt to
represent a comprehensive empirical analysis on the political economy of
settlements in investment arbitration. Such endeavour would entail a significantly
more ambitious and broader research effort. The analysis presented in this study is
<www.transnational-dispute-management.com/article.asp?key=110>; Todd J Weiler and Freya Baetens (eds), New
Directions in International Economic Law. In Memorium Thomas Wa
¨lde (Martinus Nijhoff 2011).
7
For a more detailed discussion on prevention of international investor–State disputes and investor–State conflict
management, see Roberto Echandi, ‘Complementing Investor–State Dispute Resolution: A Conceptual Framework
for Investor-State Conflict Management’ in Roberto Echandi and Pierre Sauve
´(eds), Prospects in International
Investment Law and Policy (CUP 2013) 270.
8
ICSID Convention (n 5).
9
ICSID Arbitration (Additional Facility) Rules (‘Additional Facility Rules’) (April 2006).
10
Reference to terms such as ICSID arbitration, arbitration disputes and proceedings includ es proceedings
registered under both the ICSID Convention and ICSID Additional Facility Rules. The research focused on disputes
registered during 1970–2012 and settlements to those disputes notified until May 2013. Disputes administered by the
ICSID Secretariat under the UNCITRA L Rules or other ad hoc dispute settlement provisions were not covered under
this study.
11
For purposes of this study, a dispute submitted to arbitration is considered to be ‘settled’ if the parties to the
dispute explicitly or implicitly agree to discontinue the dispute before the tribunal renders an award on merits of the
case. Thus, disputes that were settled after an award on the merits—that is, in the cont ext of an annulment
proceeding—have not been taken into consideration in this analysis.
42 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
based on publicly available data. The authors worked on the basis of the
procedural information describing the arbitration proceedings available either on
the ICSID website,
12
‘News from ICSID’,
13
or in the ICSID Review. In addition,
the identification of arbitration disputes submitted pursuant an International
Investment Agreement (IIA) was based on UNCTAD’s database of Treaty-based
Investor–State Dispute Settlement (ISDS) cases.
14
Thus, this study represents
only a very modest first step toward unveiling some factual trends in the evolution
of ICSID arbitration settlements during the period under study.
In addition to this introduction, this study includes three additional sections.
Section II starts by describing the different legal grounds and situations under which
an ICSID arbitration proceeding can be discontinued.
15
This section also examines
some empirical trends regarding the different kinds of discontinuance in ICSID
arbitration proceedings. Section III focuses on the empirical evidence of settlements
in ICSID arbitration disputes. The section provides an overview of general trends
regarding the frequency and types of settlements, the legal grounds invoked by the
parties when notifying a settlement, the timing of the settlements within the
arbitration proceedings, the settlements by economic sector, and the countries
involved in settlements. Last but not least, Section IV presents a summary of the key
findings of the research and some final reflections and conclusions.
II. THE DISCONTINUANCE OF ICSID ARBITRATION
DISPUTES
Over the past decade, it has been acknowledged that a significant percentage of
investment arbitration disputes are discontinued before an award on the merits is
rendered.
16
Table 1 provides more specific data in this respect. It shows that,
between 1970 and 2012, of the 410 arbitration disputes that were registered at
ICSID under the ICSID Convention and Additional Facility Rules, 104 disputes
were discontinued before any award on the merits had been rendered. (This
represents over 25 percent of the total number of disputes registered at ICSID
between 1970 and 2012.) Further, Table 1 also shows that the absolute number of
proceedings discontinued increased dramatically during the period. This trend
probably reflects a parallel trend with the significant increase in the arbitration
caseload that ICSID has experienced since the early 2000s.
It should not be a surprise that, at first sight, the relative share of discontinued
arbitration disputes out of the total number of arbitration disputes registered has
tended to decrease in each subsequent decade. Indeed, the high percentage of
proceedings discontinued during the 1970s and 1980s could artificially provide a
distorted picture given the limited number of disputes submitted to ICSID during
those years.
12
International Center for Settlement of Investment Disputes <https://icsid.worldbank.org/ICSID/Index.jsp>
accessed 1 June 2013.
13
ICSID, ‘News from ICSID’ <https://icsid.worldbank.org/ICSID/FrontServlet?requestType=ICSIDNewsLetters
RH&actionVal=ViewNewsLetters>accessed 1 June 2013.
14
UNCTAD, ‘Database of Treaty-based Investor–State Dispute Settlement Cases (pending and concluded)’
<http://iiadbcases.unctad.org/>accessed 1 June 2013 (‘UNCTAD’s treaty-based ISDS database’).
15
Pursuant to the ICSID Rules of Procedure for Arbitration Proceedings (‘ICSID Arbitration Rules’) (Apr il 2006),
ICSID’s Administrative and Financial Regulations (‘Administrative and Financial Regulations’) (April 2006) and
Schedule C, Additional Facility Rules (n 9).
16
Clodfelter (n 6).
WINTER 2014 Can International Investor–State Disputes be Prevented? 43
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
A. The Different Grounds for Discontinuance of Disputes
Arbitration disputes at ICSID can be discontinued in different ways. In this
regard, ICSID Arbitration Rules, ICSID Arbitration (Additional Facility) Rules
and the Administrative and Financial Regulations
17
envisage four different
situations in which an arbitration proceeding can be discontinued, each dealing
with different factual situations.
18
First, proceedings may be discontinued upon
agreement by the parties to the dispute.
19
Second, arbitration proceedings can be
discontinued at the request of a party—subject to no objection being raised by
the other party to the dispute.
20
Third, discontinuance can result from the
parties’ failure to act during an extended period to maintain the ongoing
arbitration proceeding.
21
Fourth, arbitration proceedings can be discontinued as
a result of the parties’ failure to pay the required advances to finance the
proceedings.
22
The research undertaken by this study shows that practically all of
the arbitration disputes discontinued at ICSID have been disputes governed by
the ICSID Convention rather than the Additional Facility Rules. Thus, the
subsections below focus particularly on the explanation of the arbitration rules
applicable to the different scenarios of discontinuance governed by the ICSID
Convention—although the applicable rules under the Additional Facility are quite
similar.
Table 1. ICSID Investor–State Arbitration Disputes 1970–2012: Number of Total Disputes
Discontinued before an Arbitration Award on the Merits
Period Total Disputes
Registered
Total Disputes
Discontinued
Percentage
of Total
1970–79 9 6 67
1980–89 15 7 47
1990–99 42 15 36
2000–09 233 67 29
2010–12 111 9 8
Total 410 104 25
Source: Authors’ work based on ICSID data.
17
According to the Additional Facility Rules (n 15) art 5, the Administrative and Financial Regulations (n 15)
regs 14–16, 22–30, 34(1) shall apply mutatis mutandis in respect of arbitration proceedings under the Additional
Facility.
18
There may be a fifth situation under Article 44 of the ICSID Convention under which a dispute may be
discontinued. There is a single case I&I Beheer v Venezuela, ARB/05/4 discontinued under this Article. Article 44
provides: ‘Any arbitration proceeding shall be conducted in accordance with the provisions of this Section and, except
as the parties otherwise agree, in accordance with the Arbitration Rules in effect on the date on which the par ties
consented to arbitration. If any question of procedure ari ses which is not covered by this Section or the Arbitration
Rules or any rules agreed by the parties, the Tribunal shall decide the question.’
19
Pursuant to ICSID Arbitration Rules (n 15) r 43 and the Additional Facility Rules (n 15) art 49.
20
Pursuant to ICSID Arbitration Rules (n 15) r 44 and the Additional Facility Rules (n 15) art 50.
21
Pursuant to ICSID Arbitration Rules (n 15) r 45 and the Additional Facility Rules (n 15) art 51.
22
Pursuant to the Administrative and Financial Regulations (n 15) reg 14(3)(d).
44 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
(i) Discontinuance upon agreement
First, arbitration disputes can be discontinued upon agreement by the parties pursuant
to Rule 43 of the ICSID Arbitration Rules. Rule 43 provides:
Settlement and Discontinuance
(1) If, before the award is rendered, the parties agree on a settlement of the dispute or
otherwise to discontinue the proceeding, the Tribunal, or the Secretary-General if
the Tribunal has not yet been constituted, shall, at their written request, in an order
take note of the discontinuance of the proceeding.
(2) If the parties file with the Secretary-General the full and signed text of their
settlement and in writing request the Tribunal to embody such settlement in an
award, the Tribunal may record the settlement in the form of its award.
Rule 43(1) addresses the situation in which both parties to the dispute agree, for
any reason, to discontinue the dispute. In this situation, both parties must submit
a written request to the tribunal to discontinue the proceedings. No reasons need
to be stated, and the tribunal takes note of the discontinuance in an order. Such
order has the effect of definitively discontinuing the arbitration proceedings in that
they cannot be resumed later, even if one of the parties is interested in doing so.
Thus, if the underlying dispute persisted, after the order of discontinuance was
rendered by the Tribunal, the claimant would have to resubmit a claim to
arbitration from the start. Further, being just a procedural order, discontinuance
under Rule 43(1) is not an award, and thus, does not have any res judicata effect
on the matter of the dispute.
ICSID Arbitration Rule 43(2) addresses the situation in which the parties not
only reach an amicable resolution to the dispute, but also agree to request that the
tribunal incorporate their settlement agreement into an award. In this situation,
the settlement acquires the same legal effect as any other arbitration award.
According to Articles 53 and 54 of the ICSID Convention, the award shall not
only be binding on the parties and shall not be subject to any appeal or to any
other remedy except those provided for in the ICSID Convention itself, but also,
each ICSID Contracting State—even those not participating in the dispute—shall
recognize the award as binding and enforce the pecuniary obligations imposed by
that award ‘as if it were a final judgment of a court in that State’.
23
(ii) Discontinuance at request of a party
Second, arbitral proceedings can be discontinued pursuant to ICSID Arbitration
Rule 44, which addresses the situation where only one of the parties to the dispute
formally requests the discontinuance of the proceeding. Rule 44 provides:
Rule 44
Discontinuance at Request of a Party
If a party requests the discontinuance of the proceeding, the Tribunal, or the Secretary-
General if the Tribunal has not yet been constituted, shall in an order fix a time limit
within which the other party may state whether it opposes the discontinuance. If no
objection is made in writing within the time limit, the other party shall be deemed to have
acquiesced in the discontinuance and the Tribunal, or if appropriate the Secretary-
General, shall in an order take note of the discontinuance of the proceeding. If objection
is made, the proceeding shall continue. (emphases added)
23
ICSID Convention (n 5) art 54.
WINTER 2014 Can International Investor–State Disputes be Prevented? 45
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
As the text stipulates, for a proceeding to be discontinued pursuant to Rule 44,
both parties to the dispute have to agree—at least implicitly—to terminate the
arbitration process. This provision makes it clear that the lack of objection by one
party to the request for discontinuance submitted by another party is deemed to
represent the expression of agreement to end the proceedings.
Therefore, although Rule 44 refers to a request formally submitted by just one
party, in the end, the rule represents a procedural avenue to materialize an underlying
implicit agreement by the parties to terminate the dispute. The implicit agreement
may result from a situation where the claimant drops its original claims and the
respondent happily accepts such outcome or a situation where the parties mutually
reach a settlement to the dispute, and as a consequence or as part of that deal,
the claimant submits the request to discontinue proceedings under Rule 44.
24
(iii) Discontinuance for failure of the parties to act
Discontinuance can also result from the parties’ failure to maintain the ongoing
arbitration proceeding during an extended period. This situation is envisaged in
ICSID Arbitration Rule 45, which provides:
Discontinuance for Failure of the Parties to Act
If the parties fail to take any steps in the proceeding during six consecutive months or
such period as they may agree with the approval of the Tribunal, or of the Secretary-
General if the Tribunal has not yet been constituted, they shall be deemed to have
discontinued the proceedings and the Tribunal, or if appropriate the Secretary-General,
shall, after notice to the parties, in an order take note of the discontinuance.
The promotion of judicial economy seems to be the main rationale for Rule 45, as
it ascribes a legal effect—the discontinuance of the arbitration proceedings—to the
lack of action by the parties after a given period. The second sentence of this rule
envisages the possibility of the parties requesting a ‘time off ’ and agreeing with the
tribunal to stay the arbitration process for a given period. Such a request may
indicate the existence of ongoing efforts by the parties to reach a settlement to the
dispute. In such a situation, Rule 45 may also represent a procedural avenue for
the parties to try to reach an agreement within the timeframes agreed and thus
keep the momentum of the negotiations.
(iv) Discontinuance for lack of payment of advances
The fourth ground for discontinuing an ICSID Convention arbitration is the
failure of the parties to pay the required advances to finance the proceedings
pursuant to Administrative and Financial Regulation 14(3)(d).
25
As in any other
international arbitration, the costs of arbitral proceedings at ICSID—such as
arbitrators’ fees, travel expenses and logistical and administrative charges—must
24
It is possible to envisage a situation under which, as part of the negotiation of a settlement between a foreign
investor and the government of the host State, the latter requests the investor to be the par ty requesting the
discontinuance of the dispute. An investor’s claim against a sovereign on the basis of an alleged violation of an
obligation under international law is a serious matter for many governments. Thus, in the context of a settlement, it is
not difficult to envisage a situation in which a government that is keen to ‘save face’, includes as part of the
negotiation the requirement that the investor solicit a ‘unilateral’ discontinuance of the dispute. Such a concession
may also facilitate the government to politically sell the settlement domestically. This point is further developed in
Section III.C below.
25
Administrative and Financial Regulations (n 15) reg 14(3)(e) applies to the discontinuance of annulment
proceedings, a type of discontinuance that falls outside the scope of this study.
46 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
be covered by the parties to the dispute. In order to enable the ICSID Secretariat
to pay for these expenses, and to enable the proceedings to move forward,
Regulation 14(3)(d) establishes the obligation and procedure for parties to make
those payments.
26
In relevant part, 14(3)(d) provides:
Direct Costs of Individual Proceedings
(3) In order to enable the Centre to make the payments provided for in paragraph (2), as
well as to incur other direct expenses in connection with a proceeding ...:
...(d) ...in connection with every arbitration proceeding unless a different division is
provided for in the Arbitration Rules or is decided by the parties or the Tribunal, each party
shall pay one half of each advance or supplemental charge, without prejudice to the final
decision on the payment of the cost of an arbitration proceeding to be made by the Tribunal
pursuant to Article 61(2) of the Convention. All advances and charges shall be payable, at
the place and in the currencies specified by the Secretary-General, as soon as a request for
payment is made by him. If the amounts requested are not paid in full within 30 days, then the
Secretary-General shall inform both parties of the default and give an opportunity to either of
them to make the required payment. At any time 15 days after such information is sent by
the Secretary-General, he may move that the Commission or Tribunal stay the proceeding,
if by the date of such motion any part of the required payment is still outstanding. If any
proceeding is stayed for non-payment for a consecutive period in excess of six months, the Secretary-
General may, after notice to and as far as possible in consultation with the parties, move that the
competent body discontinue the proceeding ...(emphases added)
It is the investor—rather than the government of the host State—that almost
always requests the initiation of the arbitration process. As a claimant, the investor
tends to be more interested in moving the arbitration process forward. Thus,
although Regulation (14)(3)(d) refers to both parties, in practice, one could argue
that in most cases it will apply to situations in which the investor as complainant
decides not to continue financing the arbitration process, and thus abandons the
claim. Indeed, if governments from host States—which are respondents in most
cases—were able to paralyse arbitrations instituted against them simply by abstaining
from paying the advances to finance the proceedings, one could imagine very few
ICSID arbitrations—if any at all—going forward. Thus, the system is designed in
such a way that if the host State does not contribute to the arbitration proceedings,
the claimant can fund the case. Once the decision is rendered, the award on costs can
factor in such a variable and compensate the claimant accordingly.
In sum, it could be argued that the application of Regulation 14(3)(d) as a basis
for the discontinuance of ICSID arbitrations could, to a great extent, reflect the
frequency with which investors unilaterally drop their claims, even in the absence
of any settlement offered by the respondent State. This conclusion is derived from
the premise that, had an investor been offered a satisfactory deal to settle
the dispute, there would have instead been a settlement under either Rule 43 or
Rule 44.
27
26
Under the Additional Facility Rules (n 15) ar t 5, this regulation also applies to disputes registered under the
Additional Facility Rules.
27
It could be argued that cases where investors unilaterally drop their claims could also lead to the discontinuance
of the arbitration proceedings pursuant to Rule 45. That may indeed be possible. However, if Rule 45 had to be
invoked by the Tribunal to discontinue an arbitration proceeding, it is likely that at least some payment of advances
pursuant to Regulation 14(3)(d) would have had to be previously made to put the proceedings in motion in the first
place. The more resources an investor devotes to finance the dispute, the higher the oppor tunity cost it may suffer in
dropping the claims without a settlement. Thus, one could argue that discontinuance pursuant to Regulation 14(3)(d)
may be a better proxy to reflect investors unilaterally giving up their claim than termination pursuant to Rule 45.
WINTER 2014 Can International Investor–State Disputes be Prevented? 47
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
B. Trends from Empirical Evidence
As shown in Figure 1, of the 104 arbitration disputes that were discontinued at
ICSID between 1970 and 2012, 60 percent ended pursuant to Rule 43 and 28
percent were terminated pursuant to Rule 44. Regardless of their peculiarities,
these two rules have something in common. Rules 43 and 44 are provisions that
address factual situations in which parties to the dispute have reached some level
of agreement—explicit or implicit—to discontinue the dispute. Thus, the figures
above suggest that out of the total number of arbitration disputes discontinued
during the period—including the single dispute settled pursuant to Article 49 of
the Arbitration Additional Facility Rules
28
—88 percent were based on explicit or
implicit agreements. Section III of this study is devoted to the analysis of ICSID
settlements—and thus, it will focus on the application of Rules 43 and 44. In this
section, we will centre our analysis on the remaining 12 percent of the disputes
that were discontinued pursuant to Rule 45, and Regulation 14(3)(d).
As illustrated in Figure 1, in more than 42 years of arbitration practice under
ICSID up to December 2012, only one dispute had been discontinued for failure
of the parties to act pursuant to Rule 45. The fact that only a single dispute has
been discontinued pursuant to Rule 45 clearly indicates that, in practice, investor–
State arbitrations do not terminate simply because of lack of engagement by the
parties. This finding tends to confirm the conventional view that investor–State
international arbitration may often entail significant economic and political costs
for both investors and States. Thus, it is not likely that the parties to the dispute
simply let it fade away. Further, as stated earlier, it could be argued that the
application of Rule 45 may be the result of a ‘time off’ period requested by the
parties to try to explore a settlement, and during that period, particular
circumstances led either the investor to desist from the claim, or the parties
chose to lower the profile of the dispute and mutually find a solution to their
problem. In any case, the data suggests that such a scenario has never actually
occurred during the period under study.
Table 2 shows that between 1970 and 2012, only nine arbitration proceedings at
ICSID were discontinued because of lack of payment of advances pursuant to
Regulation 14(3)(d).
The data above demonstrates that even though it is uncommon for investors to
drop their claims unilaterally—with nothing in return—once they have initiated the
arbitration process, situations like this do, on occasion, occur. From the empirical
data available, what else can we know about this uncommon, yet existent,
behaviour of the claimants? Our research sheds some light in this regard.
Out of the nine arbitration proceedings discontinued on the basis of Regulation
14(3)(d) between 1970 and 2012, three were disputes based exclusively on a
contract (‘contract-based disputes’), and six were arbitrations based on a treaty
(‘treaty-based disputes’).
29
Further, with just one exception in the 1990s, all
arbitration disputes terminated pursuant to Regulation 14(3)(d) took place
between 2000 and 2009. There were none during the 1970s and 1980s.
28
While two disputes were discontinued under Ar ticle 49, only one was settled.
29
Authors’ work based on ICSID data and UNCTAD’s Treaty-based Investor–State Dispute Settlement database
(n 14).
48 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
A more interesting trend may be unveiled by observing the procedural stage at
which these disputes were discontinued pursuant to Regulation 14(3)(d). Six out
of the nine cases—almost 67 percent—were terminated at an early procedural
stage, that is, between the registration of the dispute and the time the tribunal
rendered its award on jurisdiction, and the remaining three were settled after the
jurisdiction was upheld, but before the hearing on the merits was completed.
These facts may suggest some preliminary inferences.
First, the fact that in over 45 years of ICSID arbitration practice only seven
claims were discontinued pursuant to Regulation 14(3)(d) before an award on the
merits shows that in most cases, when investors decide to submit a dispute to
arbitration, they will go all the way until they get an award or reach a satisfactory
settlement to the dispute. This trend should be no surprise as, given the economic
and political cost involved in bringing a dispute against the host government,
investor–State arbitration is often the ‘measure of last resort’ selected by investors
to redress their grievances.
Table 2. Discontinuance of ICSID Arbitration Proceedings 1970–2012, by Applicable
Arbitration Rule
Period Rule 43 Rule 44 Rule 45 Regulation
14 (3)(d)
Article 49.1
(Additional
Facility Rules)
Article 44
ICSID
Convention
Total
1970–79 4 2 0 0 0 0 6
1980–89 5 2 0 0 0 0 7
1990–99 8 6 0 1 0 0 15
2000–09 41 16 0 8 1 1 67
2010–12 4 3 1 0 1 0 9
Total 62 29 1 9 2 1104
Source: Authors’ work based on ICSID data.
60%
29%
1% 9% 2% 1%
Rule 43 Rule 44
Rule 45 Regulaon 14(3) (d)
Arcle 49.1 Addl. Facility Rules Arcle 44 ICSID Convenon
Figure 1. Discontinuance of Proceedings by Applicable Arbitration Rule, 1970–2012.
Source: Authors’ work based on ICSID data.
WINTER 2014 Can International Investor–State Disputes be Prevented? 49
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
Second, this data also shows that when investors dropped their claims pursuant
to Regulation 14(3)(d), they tended to do so at an early stage of the arbitration
process. A possible explanation for this trend may be that, in some cases, investors
may use the registration of a dispute before ICSID as part of a strategy vis-a
`-vis
host governments to push for settlement negotiations. It has been recognized that
the prospect of investor–State arbitration may be a potential factor inducing
negotiated settlements of investor–State disputes.
30
We agree. However, such a
proposition would still fail to explain why after registering a dispute in ICSID, an
investor may subsequently drop the claim in the absence of a settlement. One
possible explanation could be that the investor had failed to properly assess the
costs of arbitration proceedings, and after being requested to deposit the cash
advances, realized that the cost–benefit assessment of pursuing the claim did not
justify continuation of the proceedings. Another possible explanation could be that
claimants may initiate arbitration despite the weakness of their case, hoping that
the prospect of being exposed to an international arbitration dispute may induce
the host government to come to the bargaining table and seriously try to reach a
settlement. By this logic, once claimants discover that the initiation of the dispute
has not rendered the expected result, it may make more sense for them to stop
spending resources and let the ICSID Secretariat apply Regulation 14(3)(d) to
discontinue the arbitration proceeding. Indeed, in this scenario, investors may no
longer incur additional costs in financing an arbitration process which is unlikely
to lead to a favourable outcome anyway.
III. SETTLEMENTS OF ICSID ARBITRATION
DISPUTES
This section analyses the empirical evidence related to settlements. In this study, a
settlement is understood as an agreement between the parties, either explicit
pursuant to Rule 43, or implicit pursuant to Rule 44, to discontinue arbitration
proceedings before the tribunal has rendered any award on the merits. Here we
attempt to examine the empirical evidence available from different points of view
in order to explore whether such data can serve as a basis for inferring behavioural
patterns of the parties to a dispute when they are attempting to settle an
arbitration proceeding.
A. General Trends in Settlements
As explained in Section II above, out of the total 104 cases discontinued in ICSID
between 1970 and 2012, the overwhelming majority—92 cases, representing
88 percent of the total disputes discontinued—were settlements.
31
As shown in
Table 3, this figure is equivalent to 22 percent of the total arbitration disputes
registered at ICSID during the period of study.
The historical trend of settlements illustrated by Table 3 shows that the absolute
number of settlements has tended to increase over the period under study. While
there were only 6 and 7 settlements respectively, this figure doubled in the 1990s,
30
Salacuse (n 4).
31
This figure includes one dispute settled pursuant to the Additional Facility Rules (n 15) art 49.
50 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
when 14 settlements were reached. Further, during the first decade of the 2000s,
the number of settlements multiplied again almost by a factor of four compared to
the previous decade. During 2000–09, 57 disputes were amicably settled before an
award on merits was rendered.
From a relative perspective, the percentage of disputes amicably settled out of
the total disputes registered in each decade has been decreasing over time. As
Table 3 demonstrates, while in the 1970s the six settled cases represented 67
percent of the total disputes registered at ICSID, in the 1980s the percentage
decreased to 47 percent, in the 1990s to 33 percent, and in the first decade of the
2000s it reached 24 percent. While it is still too early to assess the 2010s, out of
111 new disputes registered at ICSID during 2010–12, eight have already been
settled, representing 7 percent of the total so far.
However, the relative share of settlements cannot lead to the conclusion that
settlement activity in ICSID is decreasing. A proper assessment of the data has to
take into consideration that the high percentages of settlements reported during
the 1970s and 1980s distort the picture, given the limited number of disputes
submitted to ICSID during those two decades. Indeed, from the 1980s to the
1990s, the caseload at ICSID more than doubled, and during the first decade of
the 2000s the number of arbitration disputes at ICSID multiplied by a factor of
nine compared to the 1970s and 1980s. This fact explains why, despite the
significant increase in the absolute number of settlements during the period of
study, their relative weight in the total caseload registered in each decade has
tended to decrease. It could be argued that as the number of disputes registered in
ICSID tends to stabilize over time, the number and rate of disputes settled could
also follow a similar pattern—however, it is still too early to examine this
proposition from an empirical point of view.
B. Treaty-based versus Contract-based Disputes
Investment has been one of the most dynamic areas in international economic law
over the past 20 years. This is not only because of the multiplication of the
number of IIAs negotiated, but also because of the significant increase in the
number of international investor–State arbitrations requiring the interpretation of
IIAs. Considering that the first ICSID dispute invoking an IIA was registered in
Table 3. Number of Settlements Reached before an Arbitration Award on the Merits,
1970–2012
Period Total Disputes
Registered
Total Disputes
Settled
Percentage
of Total
1970–79 9 6 67%
1980–89 15 7 47%
1990–99 42 14 33%
2000–09 233 57 24%
2010–12 111 8 7%
Total 410 92 22%
Source: Authors’ work based on ICSID data.
WINTER 2014 Can International Investor–State Disputes be Prevented? 51
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
1987,
32
it is easy to forget that the international investment regime is relatively
young and still in development. Experts have argued that one of the barriers
preventing investors and host governments from finding amicable solutions to their
disputes frequently stems from the lack of clarity and the evolving nature of
international investment law.
This research attempts to elucidate whether any pattern arises when comparing
settlements in contract-based and treaty-based disputes. From the outset, it should
be pointed out that the differentiation between contract-based and treaty-based
disputes is far from easy and can be somewhat artificial. Often investors may base
their claims on both IIAs and contracts. Further, the violation of a contract may
also amount to violation of an IIA. Thus, with this important caveat in mind, our
research attempted to explore whether any pattern could be discerned on the basis
of data available. UNCTAD has prepared a database identifying those disputes in
which IIAs have been invoked in investor–State arbitration. Combining that
database with information available on ICSID disputes, we have made an
approximate differentiation between contract-based and treaty-based disputes
which were settled in ICSID before an award on the merits between 1970–2010.
Table 4
33
illustrates these findings. Of the 78 contract-based disputes registered at
ICSID in that period, 28 of them—equivalent to 36 percent of the total—were
settled. This figure is higher than the average rate of settlements in ICSID for all
kinds of disputes—22 percent—suggesting that contract-based disputes have a
higher-than-average chance of being settled amicably by the parties before a
tribunal renders an award. However, it could be argued that such a relatively high
settlement rate has to be read with caution not only because of the caveats
previously mentioned, but also due to the apparent low number of contract-based
disputes registered during the period—78 in total—relative to the total ICSID
arbitration caseload.
The relatively high percentage of settlements in contract-based disputes
contrasts with the rate of settlements of treaty-based disputes. Between 1970–
2010, 22 percent of all treaty-based disputes were settled. This figure is closer to
the average rate of settlements of the total caseload in ICSID during the period
despite being much lesser than settlements of contract-based disputes. The
contrast between the rate of settlements in contract-based and treaty-based
disputes is illustrated graphically in Figure 2.
Figure 2 clearly shows that, in proportion, settlements in treaty-based disputes
seem to be less frequent than in contract-based disputes. This fact tends to
support the opinion of a number of experts who argue that one of the barriers
preventing the parties to investor–State arbitration from reaching more agreements
is the lack of coherence in the evolving investor–State jurisprudence, which makes
it very difficult for the parties to assess the strength of their positions, anticipate
possible outcomes to the arbitration and then negotiate in the ‘shadow of the
law’.
34
Parties may be unable to assess the relative strengths and weaknesses of
32
Asian Agricultural Products Limited v Democratic Socialist Republic of Sri Lanka , ICSID Case No ARB/87/3, Award
(27 June 1990).
33
This table is based on UNCTAD’s Treaty-based ISDS database (n 14), which covers cases between 1970 and
2010.
34
On the issue of coherence and consistency in international investment law, see Christoph Schreuer, ‘Coherence
and Consistency in International Investment Law’ in Echandi and Sauve
´(n 7) 391; Peter Muchlinski, ‘Towards a
52 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
their own and opponent’s cases.
35
On the other hand, contracts contain more
precise and specific clauses drafted by the parties themselves, making it easier for
them to negotiate with greater certainty of their respective positions.
C. Choice of Arbitration Rule
As explained in Section II of this study, based on the rule invoked, settlements in
ICSID may take at least three forms. First, there may be an explicit agreement
between the parties to settle or discontinue the dispute pursuant to Rule 43(1).
36%
22%
Contract-based disputes Treaty-based disputes
Figure 2. Settlements in Treaty-based versus Settlements in Contract-based Disputes,
1970–2010. Source: Authors’ work based on ICSID data and UNCTAD’s ISDS database.
Table 4. Number of Settlements before an Arbitration Award on the Merits in Contract-
based Disputes, 1970–2010
Period Total Contract-based
Disputes Registered
Disputes
Settled
Percentage
of Total
1970–79 9 6 67%
1980–89 14 7 50%
1990–99 20 5 25%
2000–09 27 8 30%
2010–10 8 2 25%
Total 78 28 36%
Source: Authors’ work based on ICSID data and UNCTAD’s Treaty-based ISDS database.
Coherent International Investment System: Key Issues in the Refor m of International Investment Law’ in Echandi
and Sauve
´(n 7) 411.
35
Clodfelter (n 6) 41.
WINTER 2014 Can International Investor–State Disputes be Prevented? 53
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
Second, the parties may request that the tribunal embody their settlement in an
award pursuant to Rule 43(2). Third, the agreement to discontinue arbitration
proceedings may be implicit, with one party requesting the discontinuance
pursuant to Rule 44, and the other implicitly agreeing to such discontinuance.
Which kind of settlement was the most common in ICSID during the period?
What inferences—if any—could be drawn from any trend derived from the
evidence? These are the two questions this subsection purports to address.
As shown in Table 2, of the 92 disputes settled at ICSID between 1970 and
2012, 63 were a result of explicit agreement to discontinue the proceedings,
36
and
only 29 resulted from ‘implicit’ agreements to discontinue the proceedings
pursuant to Rule 44. Thus, in principle, the number of explicit settlements during
the period is double the number of ‘implicit’ agreements to discontinue arbitration
proceedings. As shown by Figure 3, the most common type of settlement in
ICSID—representing almost half of the total—are those that are reached pursuant
to Rule 43(1). Only in 18 percent of settlements did the parties request the
tribunal to embody their agreement in an award pursuant to Rule 43(2).
The importance of assessing which rule is invoked for discontinuance of the
dispute stems from the fact that each rule may tell a different story as to how the
dispute was settled and may be an additional element contributing to under-
standing the political economy underlying the settlement of arbitration disputes.
Although further research would be needed to test such propositions, in the
following paragraphs, we suggest some preliminary hypotheses on the political
economy of ICSID settlements depending on the rule invoked to discontinue the
arbitration proceeding.
49%
18%
32%
1%
Rule 43.1 Rule 43.2 Rule 44 Arcle 49.1 Addl. Facility Rules
Figure 3. Settlements by Applicable Rule, 1970–2012. Source: Authors’ work based on
ICSID data.
36
Sixty-two arbitration proceedings were settled pursuant to Rule 43.1 and Rule 43.2, and one pursuant to Article
49.1 of the Additional Facility Rules (n 15).
54 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
During the period under study, Rule 44 was invoked to request the
discontinuance of 29 arbitration disputes. In 27 of those 29 settlements,
the request to discontinue the proceeding came from the claimant (ie generally
the investor). This raises the question: what are the political economy consider-
ations leading an investor to request the discontinuance of the dispute [after
having advanced payments pursuant to Regulation 14(3)(d)]? It could be argued
that an investor is unlikely to request the discontinuance of arbitration proceedings
pursuant to Rule 44 without a clear change in the circumstances that motivated
the registration of the dispute in the first place.
It is difficult to speculate what such circumstances might be. It may be that the
investor no longer considers the dispute worth pursuing—or that there was a change
of policy within the board of the investing company. It may also be possible that,
although Rule 44 theoretically refers to a unilateral request for discontinuance by just
one party, de facto, it is an avenue to implement an underlying settlement of the
dispute by the parties. For many governments, reaching a ‘settlement’ with a foreign
investor who has accused a State of disregarding its international obligations may not
be politically easy. Thus, even in situations in which government officials may
understand that an eventual arbitration award will confirm most of the investor’s
claims, governments may need to ‘save face’ and portray the settlement to their
domestic constituents as a situation in which the investor ‘gave up’ its allegations.
Our research has shown that the most common type of settlements in ICSID
arbitration disputes between 1970 and 2012, 45 in total, are those made pursuant
to Rule 43(1). These are explicit agreements reached among the parties to
discontinue arbitration proceedings without requesting that the tribunal embody
their agreement into an award. This trend suggests that in most cases where a
settlement is reached, governments do not have a problem recognizing that a
negotiation with the investor has taken place and that a ‘mutually accepted
solution’ to the dispute has been reached.
37
Further, in addition to the 45
arbitration disputes discontinued pursuant to Rule 43(1), as shown by Table 5, 17
disputes were discontinued pursuant to Rule 43(2) during the period under study.
Under Rule 43(2) the investor may request the incorporation of a settlement
into an award of the tribunal, making it enforceable under the ICSID Convention.
In the absence of incorporation of the settlement agreement into the award, such
advantage would not operate. This reasoning may suggest that, in certain
circumstances, an investor with superior bargaining power would ideally like to
‘seal’ the settlement agreement in an award. On the other hand, as a condition for
requesting to discontinue the arbitration proceedings, the investor and the
government may have managed to implement the actions agreed between them
in order to settle the dispute, and thus it may no longer be necessary to embody
the settlement in an award. Thus, whether an arbitration proceeding was dis-
continued pursuant to Rule 43(1) or Rule 43(2) does not say much about the
relative bargaining power positions of the parties involved in the settlement
negotiations.
37
It could be argued that changes in government may facilitate the settlement of pending disputes in situations
where new authorities want to differentiate themselves from previous administrations having a contentious record with
foreign investors. In these cases, new regimes may not be so interested in ‘saving face’, but rather in sending a positive
signal to the international investment community of the ‘positive’ attitudes of the new government to foreign
investment.
WINTER 2014 Can International Investor–State Disputes be Prevented? 55
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
D. Timing of Settlements
One of the central questions that this research attempted to address was whether
the settlements in arbitration disputes took place at an early or later procedural
stage. The reason for examining this variable is that it could provide an objective
indication of the minimum number of disputes that might have been prevented
from escalating into international arbitration if the parties had a proper outlet to
manage their conflict. In this regard, our hypothesis is that if settlements are taking
place so early that the parties have not used the arbitration process to assess the
strength of their respective claims, then the need for a tribunal to elucidate on a
disagreement of law or fact would be questionable. Let us develop this point
further.
In principle, one may assume that the attitudes of the parties toward engaging in
settlement negotiations may be related to the perception of their chances of having
the tribunal accept most of their respective claims. For instance, if the authorities
of the respondent government are convinced that the investor’s claims are weak
and have very little chance to succeed, the incentive for government officials to
compromise and assume the political responsibility of settling the dispute would
be relatively low. Conversely, it could be assumed that the clearer the prospects of
the investor’s claims being accepted by the tribunal, the easier it will be for
government officials to get the political backing required to reach a compromise
with the investor bringing the dispute.
The arbitration process enables claimants and respondents to interact and learn
about each other’s claims and arguments vis-a
`-vis the issues of law and fact subject
to dispute. Further, as proceedings advance, the parties usually develop a better
assessment of which arguments and claims are stronger or weaker, enabling them
gradually to draw a better picture of the possible outcome of the arbitration
process. The parties’ clarity regarding the strength of each other’s claims and
arguments tends to increase as the different stages of the arbitration proceedings
advance. Conversely, the earlier it is in the process, the less able the parties are to
assess the strength of each other’s cases.
If a settlement takes place before the parties have had a chance to assess each
other’s claims and arguments—or even know them in full detail—then it follows
that such settlement was based on the parties’ own preliminary assessment of the
merits of the case. In other words, it is not the evolution of the discussion during
Table 5. ICSID Investor–State Arbitration Disputes 1970–2012: Number of Settlements
Pursuant to ICSID Arbitration Rules 43.1, 43.2 and 44
Period Rule 43.1 Rule 43.2 Rule 44
1970–79 3 1 2
1980–89 5 0 2
1990–99 3 5 6
2000–09 31 10 16
2010–12 3 1 3
Total 45 17 29
Source: Authors’ work based on ICSID data.
56 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
the arbitration process that provides the parties with better information to ponder
the cost and benefits of settling the dispute—as in early stages of the proceedings
such interaction has not yet taken place. Rather, it would seem that, even before
the arbitration process started, the parties already had enough information to
enable them to settle, and that it is the very prospect of imminent adjudication
that finally leads the parties to conclude the settlement.
To test our hypothesis, we observed the timing of settlements between the
parties to the dispute. For such purpose, arbitration proceedings were divided into
four phases. As shown in Figure 4, Phase 1 comprises the period between
registration of the dispute by the ICSID Secretariat and the time before the
discussion on jurisdiction begins. Phase 2 comprises the period during the
discussion on jurisdiction until and the award on jurisdiction, or in those
arbitration proceedings which do not have a separate phase on jurisdiction, until
the parties have exchanged their first memorials and counter memorials. Phase 3
covers the period after the award on jurisdiction—or after the exchange of
memorials and counter memorials in arbitrations without a phase on jurisdiction—
until the completion of the hearing on merits of the case. Phase 4 comprises the
period after the completion of hearing until the final award on merits is rendered.
Our research found that most of the settlements during 1970 and 2012—39
percent of the total—took place in Phase 1 of the arbitration proceedings, between
the registration of the dispute and before the discussion on jurisdiction begins.
Table 6 provides the dynamic picture of the historical evolution of settlements
during the period covered in the study. While it is evident that settlements in
Phases 1 and 2 have tended to increase, the settlements in Phase 3 have followed
the reverse trend, and have consistently become less frequent over the years. It is
worth noting that as the caseload of ICSID increased, settlements have tended to
take place earlier in the arbitration process.
The trend toward earlier settlements in the arbitration process has been
consistent in both treaty-based and contract-based disputes. In the case of treaty-
based disputes, 75 percent of the settlements took place in Phase 3 in the 1990s,
while in the 2000s, most of the settlements happened in Phase 1—41 percent of
the total.
38
This trend has continued during 2010–12, when all of the settlements
of treaty-based disputes so far have taken place in early phases.
39
Phase 1
Registration Const. Tribunal
Award on
Jurisdiction Audience
Merits Award/Findings
on liability
Phase 2 Phase 3 Phase 4
Figure 4. Arbitration Proceedings Timeline—Phases.
38
Authors’ work based on ICSID data.
39
ibid.
WINTER 2014 Can International Investor–State Disputes be Prevented? 57
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
In the case of settlements involving contract-based disputes, the trend toward
more settlements in earlier phases of the arbitration process over time is also clear.
While in the 1970s there was not a single dispute settled in Phase 1, in the 1980s
disputes settled in Phase 1 amounted to 14 percent. In the 1990s that figure climbed
to 20 percent, and in the 2000s 62 percent of the total number of contract-based
settlements occurred in Phase 1.
40
Thus, the trend for the parties to settle in early
stages of the arbitration process is clear both in treaty-based and contract-based
disputes.
In sum, empirical evidence suggests that settlements are increasingly taking
place earlier in the arbitration proceedings. What does this trend tell us—if
anything—from a political economy perspective? Why do most settlements take
place in Phase 1 and not in Phase 4 when the parties are likely to have a better
appraisal of the possible outcome of the case? This trend is particularly puzzling
considering that empirical evidence has shown that investors do not necessarily
tend to win the majority of investor-State disputes.
41
The increase in the number of early settlements may suggest various things. The
first inference is that disputes that are settled may be based on conduct that is
clearly inconsistent with the obligations of the host country vis-a
`-vis the investor.
A second inference is that, as the caseload of ICSID arbitration has expanded,
international investment law jurisprudence has started to develop, providing more
elements of judgment to the parties to assess their claims. In addition, the increase
in investor–State litigation activity in the past two decades has increased the
visibility of ICSID arbitration proceedings, thus increasing the potential reputa-
tional cost for a State of acting inconsistently with its commitments vis-a
`-vis
international investors.
Indeed, given the accountability of public officials and the visibility of an ICSID
arbitration, it is unlikely that any government representative would assume the
political cost of reaching a settlement with an investor which may entail the use of
public resources in the absence of a clear justification.
42
Such justification may
stem from the imminence of an adverse award against the State, or in cases where
Table 6. Distribution of Settlements by Procedural Phase, 1970–2012
Period Phase 1 Phase 2 Phase 3 Phase 4
1970–79 0 50 50 0
1980–89 14 43 43 0
1990–99 29 14 36 21
2000–09 44 28 26 2
2010–12 75 25 0 0
Total 1970–2012 39 28 28 4
Source: Authors’ work based on ICSID data. Values are represented as percentages. Final
percentages may not total to 100 due to rounding.
40
ibid.
41
Susan D Franck, ‘Empirically Evaluating Claims about Investment Treaty Arbitration’ (2007) 86 North Carolina
LRev1.
42
Michael Reisman, ‘International Investment Arbitration and ADR-Married but Best Living Apart’ in UNCTAD
(n 6) 26.
58 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
the outcome may not be completely foreseeable, from a clear assessment that the
political and economic cost of continuing with the dispute may be higher than that
of the settlement. The fact that settlements are tending to take place in early stages
of the process means that the assessment leading to the conclusion of a settlement
can be undertaken by the parties themselves, without the help of the tribunal.
In sum, governments may be settling in those situations where it is clear that
they may lose the case. Further, presumably, in most cases settlement negotiations
may have begun before the case is even registered, and the registration of the
dispute at ICSID becomes a clear signal of the determination of the investor to
redress his/her grievance. This fact also suggests that cases which are settled in the
early phases of the arbitration process could have been prevented from reaching
international arbitration had the space/platform to pursue an informed and
interest-based negotiation process been provided to the parties prior to registration
of the case.
E. Settlements by Economic Sector
In which sectors do settlements occur? Are there any sectors more conducive to
settlements? These are the questions this subsection attempts to answer.
During the period under study, most of the settlements were related to disputes
involving services, comprising 60 percent of the total settlements, whereas the
percentage of the settlement involving the primary sector is 24 and only 16 percent
relate to the settlement of disputes involving the manufacturing sector. The
continuing increase of settlements in the services sector can be easily observed by
Figure 5.
43
More importantly, the distribution of settlements by economic sector reveals
a picture which is strikingly similar to the distribution of ICSID arbitration
disputes by economic sector during the period under study. This trend is depicted
in Figure 6.
According to the 2013 ICSID caseload statistics,
44
of all disputes submitted to
ICSID between 1970 and 2012, 57 percent related to the tertiary sector (such as
information and communications, finance, transport, tourism, construction,
utilities and energy generation), while 30 percent related to the primary sector
(agriculture, fishing, forestry, oil & gas and mining) and only 13 percent related to
the manufacturing sector (other industry). During the same period, the distribu-
tion of settlements by economic sector tends to show a very similar trend: 60
percent of the settlements in the tertiary sector, 24 percent in the secondary sector
and 16 percent in the primary sector.
As can be clearly observed from Figure 6, the distribution of the ‘pies’ are quite
similar. This finding suggests that the rate of success of settlement in each
economic sector seems to be a reflection of the caseload associated with each
economic sector. Thus, the probability of reaching a settlement in an investment
dispute is unaffected by the economic sector involved.
43
Since the data for 2010–12 was insufficient to reveal a trend for the decade, it has been excluded from this
graph.
44
See ICSID, ‘The ICSID Caseload—Statistics 2013-1’ <https://icsid.worldbank.org/ICSID/FrontServlet#>
accessed 1 June 2013.
WINTER 2014 Can International Investor–State Disputes be Prevented? 59
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
F. The State Actors Involved in Settlements
Another way to examine the available empirical evidence on settlements of
arbitration disputes is by focusing on States involved in settlements. What are the
countries that have settled? Are there some countries that settle more than others?
This section examines whether some inferences can be drawn regarding the profile
of the countries that have settled during the period under study.
From the outset, it should be noted that a country cannot be involved in a
settlement if it has never been involved in an arbitration dispute. In this regard,
out of the 150
45
States which were Contracting States
46
of ICSID during the
period under study, 100 were involved in at least one dispute—a figure equivalent
to 68 percent of the total number of its Contracting States as of 2012 (ie 147).
This means that approximately 32 percent of ICSID Contracting States have not
yet had any ICSID arbitration experience.
0
10
20
30
40
50
60
70
1970s 1980s 1990s 2000s
Primary Secondary Terary
Figure 5. Historical Trend of Settlements by Sector. Source: Authors’ work based on
ICSID data.
57% 60%
13%
30% 24%
16%
Primary
Secondary
Terary
Primary
Secondary
Terary
Figure 6. Distribution of ICSID Arbitration Disputes by Economic Sector: Comparison
between Disputes Registered and Disputes Settled. Distribution of disputes by economic
sector, 1970–2012 (left panel). Distribution of settlements by economic sector, 1970–2012
(right panel). Source: Authors’ work based on ICSID data.
45
As of December 2012, ICSID had 147 Contracting States. This excludes Sao Tome and Principe and
Montenegro, which became Contracting States in 2013. This includes Bolivia which ceased to be a Contracting State
in November 2007, Ecuador in January 2010 and Venezuela in July 2012 <http://icsid.worldbank.org/ICSID>
accessed 1 June 2013.
46
States attain the status of ‘Contracting State’ when the ICSID Convention enters into force in their territories.
60 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
Further, of the 100 countries that have been respondents in an ICSID
arbitration dispute between 1970 and 2012, 54 have agreed to a settlement on at
least one occasion. This means that around half of the countries which faced an
arbitration dispute have had at least one settlement, while approximately the other
half has not settled a single dispute—46 countries to be exact.
Are there countries which have settled more than others? The answer depends
on whether the number of settlements is analysed on the basis of the absolute
number of settlements reached by each country or on the basis of the number of
settlements relative to the number of disputes in which the countries have been
involved.
From an absolute perspective, the country with the highest number of
settlements during the period is Argentina, with 14 settlements, followed by
Ecuador with six; Pakistan with four; Jamaica, Jordan, Venezuela, Democratic
Republic of Congo and Georgia with three settlements each; and Azerbaijan,
Bolivia, Congo, Gabon, Morocco, Nigeria and Ukraine, each with two settle-
ments. The fact that the countries with a higher number of settlements are also
countries with a high number of disputes—clearly the case of Argentina—may
quickly lead one to think that there is a correlation between the absolute number
of disputes in which a country is involved and the number of settlements that such
country may reach. However, that is not the case. In fact, there are countries
which have faced a significant number of ICSID arbitration disputes during the
period and yet have not reached a single settlement. Thus, our research attempted
to assess the number of settlements reached by countries relative to the number of
disputes registered against them.
From a relative point of view, our research shows that while there are countries
which have settled all the disputes in which they have been involved, there are
others which have not settled a single one of the cases brought against them.
47
Focusing on the countries that have settled at least one ICSID dispute, and
observing the percentage of the disputes settled out of the total disputes in which
they have been involved, we classified countries in four categories. Category ‘A’
comprises countries that have settled 100 percent of the disputes in which they
have been involved. Category ‘B’ covers countries that have settled 50 percent or
more of their disputes. Category ‘C’ includes countries that have settled 30
percent or more of their disputes, and Category ‘D’ comprises those countries that
have settled at least one dispute, but less than 30 percent of the total disputes
brought against them. Table 7 captures the settlement profile of each of the four
categories.
The 13 countries falling within Category ‘A’ together total 15 disputes settled,
representing 16 percent of the total ICSID arbitration settlements during the
period. In order to build a profile of the kind of countries that fall in this category,
it is worth noting that with the exception of Jamaica (which had three disputes but
all of which were related to the same matter), all the countries in Category ‘A’
47
From this fact, however, we cannot deduct any par ticular attitude of governments toward settling disputes. An
agreement always involves at least two parties, and thus the lack of settlements by a country may stem from various
factors, some of which may be attributed to attitudes of investors, public policy concerns, the kind of issues involved
in the disputes, etc.
WINTER 2014 Can International Investor–State Disputes be Prevented? 61
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
have had only one investment dispute brought against them during the period. In
53 percent of the cases settled, these countries have settled promptly—that is, in
Phase 1 of the arbitration process.
Category ‘B’ comprises 18 countries, which have settled 29 disputes in total,
representing 32 percent of the total number of settlements. The common feature
of all countries falling within this category is that all of them have faced at least
two disputes. In fact, only four countries have been involved in four or more cases,
the rest having faced two or three cases. It is worth noting that this category
comprises the largest number of countries and settlements.
The 11 countries comprising Category ‘C’ together total 20 disputes settled,
representing 22 percent of the total settlements. Some of the countries with the
highest number of ICSID arbitrations fall here—such as Ecuador, with 14 cases.
Thus, it appears that a heavy caseload does not necessarily prevent countries from
settling a significant share of their claims. However, once again, the common
feature among all countries falling in Category ‘C’ is that all of them have been
involved in at least three arbitration disputes.
Category ‘D’ comprises 12 countries, which together cover 28 settlements,
representing 30 percent of the total. This category comprises Argentina, which has
settled 29 percent of its total disputes and thus falls short by a single case to be
classified in Category ‘C’. Excluding Argentina, this category would be signifi-
cantly smaller, and would comprise 11 countries, adding up to only 14 of the total
number of settlements between 1970 and 2012. As in the case of the other
categories, once again, all of the countries falling within Category ‘D’ have one
common feature—they all have acted as respondents in at least four arbitration
disputes. The data shown in Table 7 seems to suggest a pattern: the higher the
number of disputes a country faces, the more difficulty it faces in maintaining a
rate of settlement of 30 percent or more of its total caseload.
Table 7 also suggests an interesting pattern regarding the countries that have a
100 percent rate of settlements. These are all countries that have faced an ICSID
dispute for the first time and a majority of the countries falling in this category
have also settled in Phase 1 of the arbitration process. These observations clearly
indicate the interest of these governments in settling their disputes promptly. They
Table 7. Profile of Each Category of Countries that Settle
Category of Countries Number of
Countries in
the Category
Percentage of
Total Countries
that have Settled
at least One
Dispute
Cumulative
Number of
Settlements
Percentage
of Total
Settlements
A. Settled 100% of disputes 13 24% 15 16%
B. Settled 50% or more 18 33% 29 32%
C. Settled 30% or more 11 20% 20 22%
D. Settled less than 30% 12 22% 28 30%
Total 54 100% 92 100%
Source: Authors’ work based on ICSID data. Some columns do not total to 100 due to
rounding.
62 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
also suggest that governments do care about the possible repercussions from their
first ICSID arbitration dispute.
Our research also attempted to find out whether there is any discernible
geographical pattern emerging from the countries that settled during the period. In
this regard, as depicted in Figure 7, the research revealed an interesting apparent
pattern: the regional distribution of arbitrations registered at ICSID to a great
extent tends to mirror the regional distribution of settlements. This fact suggests
that whatever the factors that enable governments to reach settlements, they
appear to be region-neutral, as no region is proportionally more inclined toward
settling arbitration disputes than another.
IV. CONCLUSIONS
The purpose of this study was to promote a better understanding of the profile of
settlements in investment arbitration. In particular, our research purported to
address a concrete problem on the basis of empirical evidence. That is, whether,
and to what extent, investment arbitration is dealing with disputes that should not
have escalated to the international arbitration level in the first place, as they could
reasonably have been mutually resolved by the parties at an earlier stage. To
address this question, in addition to examining other key contextual variables, our
research observed how early or late along the arbitration proceedings have
settlements taken place. Our hypothesis was that if settlements are taking place so
early in the arbitration process that the parties have not used the process to assess
the strength of their respective claims and arguments, then the real need for a
tribunal elucidating a disagreement of law or fact would be questionable. It would
then follow that in such situations, ICSID arbitration would be dealing with
disputes that, had there been a proper outlet for the parties to resolve them at the
domestic level, may have been prevented from escalating to international
adjudication.
0%
9%
9%
17%
13%
19%
4%
30%
Geographic Distribuon of Selements, 1970-2012
North America Central America & Caribbean
South America South & East Asia
Middle East & North Africa Sub-Saharan Africa
Western Europe Eastern Europe & Central Asia
5% 6%
30%
9%
10%
14%
5%
23%
Geographic Distribuon of Disputes, 1970-2012
North America Central America & Caribbean
South America South & East Asia
Middle East & North Africa Sub-Saharan Africa
Western Europe Eastern Europe & Central Asia
Figure 7. ICSID Arbitration Disputes, 1970–2012. Geographic Distribution of Disputes
(left panel) and Settlements (right panel), 1970–2012. Source: Authors’ work based on
ICSID data.
WINTER 2014 Can International Investor–State Disputes be Prevented? 63
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
This study revealed that 92 cases, representing 22 percent of the total
arbitration proceedings registered in ICSID between 1970 and 2012, were settled
before an award on the merits was rendered, as a result of an agreement between
the parties. Out of those 92 cases, 36 of them—representing 39 percent of the
total—took place during what we denominated as ‘Phase 1’ of the arbitration
proceedings—that is, between the time the arbitration was registered and the
constitution of the tribunal.
Further, observing the historical evolution of the settlements over the different
decades between 1970 and 2012 it is worth noting that settlements in Phases 1
and 2 have tended to increase, while settlements in Phase 3—that is, after the
hearing on the merits—have consistently become less frequent. Thus, over time,
the ICSID caseload has increased, and the jurisprudence interpreting international
investment law has gradually developed, while more settlements have tended to
take place earlier in the arbitration process.
As far as the economic sector is concerned, it seems that the probability of
reaching a settlement in an arbitration dispute is not affected by the fact that a
dispute relates to services, manufacturing, agriculture or extractives. Indeed, the
rate of settlement of disputes tends to mirror the caseload associated with each
economic sector. Given the increasing number of arbitration disputes involving
services, it is not surprising then that 60 percent of all the settlements were in
cases involving the tertiary sector.
Regarding the States that have been involved in settlements, we found that, out
of the 100 countries that have been respondents in at least one arbitration dispute
during the period, 54 have settled at least one dispute, while 46 have not settled
a single one. The picture of which countries tend to settle more often changes
depending on whether one counts the absolute number of settlements or rather
the percentage of disputes settled relative to the total caseload of countries. It
should be noted that between 1970 and 2012, the country with the highest
absolute number of settlements was Argentina, the country that also faced the
highest absolute number of arbitration disputes over the period. On the other
hand, countries that have settled 100 percent of their disputes are countries that
have faced only one ICSID arbitration case during the period—the majority of
which tended to settle in Phase 1 of the arbitration proceedings. This fact suggests
that at least certain governments pay attention to the possible repercussions of
their first ICSID arbitration, but as countries begin to face more than one dispute,
the probability of maintaining a ‘clean record’ of settlements starts to diminish. We
also found that countries that settle belong to all regions of the world and there is
no one particular region which is more likely to settle arbitration disputes than
another.
In sum, the findings of our research tend to support three key ideas. First, the
fact that substantial negotiation activity is already occurring in investor–State
arbitration supports the notion that the very availability of compulsory arbitration,
and in particular the prospect of facing an adverse award, increases the possibility
that respondent State governments will seriously consider amicable settlement, as
an option to resolve a dispute with affected investors. Second, the gradual
development of a body of jurisprudence interpreting international investment law
over the past two decades seems to have contributed to assisting the parties in
64 ICSID Review VOL. 29
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
better assessing the probability of getting a favourable award, and pursuing more
informed settlement negotiations.
Third, the significant—and growing—number of arbitration disputes settled at
early stages of the proceedings reveals the potential for establishing mechanisms to
enable States and investors to effectively manage their conflicts and find mutually
agreed solutions before such grievances escalate into international arbitration.
Such mechanisms could not only reduce the growing caseload, but more
importantly, could strengthen the relationships between investors and host
States by promoting problem solving without incurring the economic and political
costs of international arbitration.
WINTER 2014 Can International Investor–State Disputes be Prevented? 65
Downloaded from https://academic.oup.com/icsidreview/article-abstract/29/1/41/2356615 by Sectoral Library Rm MC-C3-220 user on 19 September 2018
... U literaturi se ukazuje na činjenicu da tužene države tokom pregovora o mirnom rešenju spora ponekad postave uslov da posle poravnanja tužilac povuče tužbu kako bi reputacija tužene ostala neokrnjena. 19 Drugim rečima, iza pozivanja na pravilo 44 često se krije poravnanje. Stranke koje ne žele da obelodane da su postigle poravnanje i nemaju interesa da arbitražni sud unese poravnanje u arbitražnu odluku, mogu se saglasiti da umesto da zajedno traže okončanje postupka na osnovu arbitražnog pravila 43, to učini tužilac na osnovu arbitražnog pravila 44, a tuženi da se sa tim saglasi. ...
Article
Full-text available
This study examines arbitration proceedings that were discontinued after being registered with the ICSID Secretariat under the ICSID Convention and ICSID Arbitration Rules or under the ICSID Additional Facility Rules between 2010 and 2019. The research focused on number of discontinued arbitration proceedings and the legal grounds invoked by the parties to discontinue the arbitration proceedings. The study revealed that settlements are the most common reason for the discontinuance of the proceedings. However, in a surprisingly small number of cases, the parties requested the arbitral tribunal to transform the settlement into an arbitral award on agreed terms. Furthermore, the number of cases discontinued due to the non-payment of deposits was equally small. Considering the grand entry of third-party funding into the field of investment arbitration, it may be predicted that the number of such cases will be even smaller in the future. The discontinuance of the proceedings due to non-activity of the parties is even less frequent. Regarding the states that have been involved in settlements, the author found that out of the 128 countries that have been respondents in at least one arbitration dispute during the period, 67 have settled at least one dispute, while 61 have not settled a single one. The picture of which countries tend to settle has changed comparing to the previous periods. The place of Serbia in this picture is not very good and should be improved. Serbia has had seven ICSID cases so far, but only one settlement. The high costs of the arbitration proceedings administered by ICSID justify investing more effort into creating a permanent mechanism which will enable settlement both in the pre-arbitration phase and after the case has been registered.
Article
La prominencia de Argentina en la historia del arreglo de diferencias relativas a inversiones internacionales (ISDS) permite un estudio de caso fundamental sobre la tensión entre las medidas estatales y la inversión extranjera directa (IED). Argentina, al igual que otros países latinoamericanos, ha adoptado un enfoque proactivo para mitigar la pandemia actual. En particular, estas decisiones de salud pública de emergencia pueden obstaculizar la IED, lo que conlleva a un aumento de las disputas de inversiones. Este artículo tiene como objetivo analizar comparativamente la crisis argentina pasada, y los casos del Tratado de Libre Comercio de América del Norte (TLCAN) relacionados con la salud publica, capitalizando sobre las lecciones aprendidas para proporcionar una guía en anticipación de las disputas de la covid-19. Para explorar este tema, una discusión sobre cuestiones jurisdiccionales y de procedimiento permite una aplicación moderna de cuestiones pasadas.
Article
In the Monetary Gold case, the International Court of Justice (ICJ) stated that it cannot exercise its jurisdiction over a question when a third state's 'legal interests would ... form the very subject-matter of the decision'. The purpose of this article is to answer the question whether this principle applies to international courts and tribunals beyond the ICJ. It begins by providing an overview of jurisprudence concerning the Monetary Gold principle. Following, it analyses the rationales suggested in jurisprudence and legal doctrine for the principle. It is argued that these suggested rationales fail to explain the nature of the principle, or constitute shaky ground for a principle of general applicability. Finally, it is argued that, regardless of the origins of the Monetary Gold principle, it can be grounded on customary international law, and on that basis applies to international courts and tribunals generally. © 2019 The Author(s). Published by Oxford University Press. All rights reserved.
Article
As a rapidly developing country, Indonesia realizes that foreign direct investment (FDI) remains necessary to fuel growth in the capital-scarce country. Yet, along with its development, Indonesia is facing more complex challenges both because of domestic issues and its problematic bilateral investment treaties (BITs) which do not provide explicit policy space for the government. The problematic implementation of the desentralisasi system, a limited rule of law and difficulties in dealing with corruption continue to reduce Indonesia’s attractiveness to foreign investors. Indonesia has decided to discontinue its old BITs and has expressed an interest in designing a new BIT model. This article attempts to suggest ways in which Indonesia can deal with its domestic challenges and to improve its new international investment agreements to create a more conducive investment climate.
Article
Full-text available
Investor-State Dispute Settlement (ISDS) clauses in international investment agreements have traditionally been based on an approach which may be termed ‘light touch regulation’ of investment protection. The avenue taken by the recently negotiated EU draft agreements, the Comprehensive Economic and Trade Agreement (CETA) and the EU-Singapore Free Trade Agreement (EUSFTA), can be described as ‘more comprehensive regulation’. Likewise, EUSFTA and CETA provide a rather detailed body of law on substantive standards for the protection of foreign investment. While this may add to the clarity and predictability of the current regime of international investment law, it may also lead to a reduced standard of protection. Compared with other agreements, EUSFTA and CETA have attempted to rebalance the protection of private property and the host state’s regulatory autonomy. In terms of the regulation of ISDS proceedings, EUSFTA and CETA preserve its principle characteristics but deliver moderate change in five areas: (1) consultation mechanisms, (2) the relationship between ISDS and domestic remedies, (3) the appointment and conduct of arbitrators, (4) cost allocation, and (5) transparency rules. This study proposes (1) further development regarding the coordination between effective domestic legal systems and ISDS and (2) the start of negotiations for the establishment of a permanent appeals mechanism in a regional or bilateral context.
ResearchGate has not been able to resolve any references for this publication.