Conflict of Law For Arbitrators
Conflict of Law For Arbitrators
Conflict of Law For Arbitrators
I. INTRODUCTION
PARTIES TO an international contract often choose arbitration because they
want a measure of certainty and predictability as well as neutrality and
effectiveness for the dispute resolution process. Besides, the parties aim at a
stable substantive law under which their disputes should be resolved. This legal
regime is the basis for the award. Since there are still differences among national
laws, the parties have a great interest in the arbitrator's determination of the law
applicable to the merits of their disputes.' Although this question is of crucial
importance, a homogeneous answer has not yet been given. Nor do resulting
arbitrations reveal any clear dominant attitude towards the application of conflict
of laws rules by international arbitrator^.^ A reason for this might be that
arbitration is secret and the awards do not have the authority of precedents,
although there may be psychological elements of persuasive authority amounting
to precedents: 'Faced with a problem one wants to know what others in similar
situations have done, and one tends to copy them'.3 Will the arbitrators apply the
conflict of laws rules pertaining to the seat of arbitration, or those of the country
which has the 'closest connection' with the disputeT4 Or will they not apply
conflict of laws rules at all? The problem can be divided into two logical steps:
the first question is whether it is necessary to apply a conflict of laws system at all,
and secondly, if the answer is positive, which system should be applied.
"
98
When the parties have agreed upon the substantive law applicable to the merits
of their dispute, it seems as if the problem of which conflict of laws rules should
be applied by the arbitrators will not arise because the arbitral tribunal can solve
the disputes by directly applying the chosen substantive law. But does this mean
that the arbitrator has to recognize the parties' autonomy without relying on any
conflict of laws rule? The answer to this question would be aMirmative if all
private international law systems accepted the freedom of the parties to choose
the applicable law without any restrictions. Although the principle of party
autonomy is basically recognized in both common law and civil law systems, the
freedom which each country permits is limited. For instance, in the English
system of private international law, a choice by parties must be 'bona fide and
legal' and a merely 'eccentric or capricious' choice of law will have no effect.5 Is it
possible to choose only some provisions of a legal system and to exclude the
others? Can the parties agree that, for example, one part of a complex joint
venture agreement is subject to the law of country A whereas another part is
Is a retroactive choice of law
subject to the regulations of country B (dkpe~age)?~
p ~ s s i b l e ?Must
~ the choice of law comply with certain formal requirements? In
order to answer these questions, the arbitrator must examine which conflict of
laws rule affects the validity of the choice of law.
Moreover, in some cases a statutory provision, for example in English law, will
apply to a contract despite the fact that the parties have chosen a foreign law
governing the merits of their ont tract.^ Thus, under English private international
law the parties cannot escape from the mandatory provisions of that legal system
with which the contract has its most substantial connection and which, for this
reason, the court would have applied if there had not been an expressed or
implied choice of law.g In general, mandatory rules comprise rules of law which
cannot be derogated from by contract, including rules regarded by a legal system
as being of such importance that they must be applied irrespective of the
99
applicable law. Since these regulations in the interest of the general public solely
achieve their objective if they are applied notwithstanding the chosen law
governing the merits of the dispute, it would violate the will of the respective
legislatures to apply only the rules of the law chosen by the parties.1 Otherwise,
the parties could circumvent those provisions which did not appeal to them.
Since arbitrators ought to have an incentive to apply mandatory rules out of a
sense of duty to the survival of international arbitration as an institution, they
must not permit recourse to arbitration to become a means for escaping
fundamental principles of public policy. By complying with compulsory
provisions representing the public policy of a state, arbitration is more likely
to be tolerated by states and their courts, and consequently to survive as a private
dispute settlement system within those states." What would be the reaction of
US courts if arbitrators did not apply the relevant American anti-trust legislation,
although in the Mitsubishi case the matter was regarded as arbitrable? Once
accepted that an arbitrator has to pay regard to mandatory rules, the logical
question which follows is with which mandatory rules an arbitrator has to
comply. Of course, he has to apply the obligatory provisions of the substantive
law chosen, even if the parties excluded them. Beyond that, conflict of laws rules
might help him to confine the scope of systems of law whose obligatory
provisions should influence the parties' dispute.12 Consequently, even if the
parties have chosen the applicable law it may be vital for arbitrators to determine
the conflict of laws rules.13
lo
"
l3
14
See e.g. Art. 34 of the Rome Convention and Art. VIII(Z)(b) of the International Monetary Fund (IMFI
Bretton Woods Convention). In respect of the latter, see the Award of the Amsterdam Grain Trade
Association cited by Berger, op. cit at p. 691, n. 267.
P. Mayer, 'Mandatory Rules of Law in International Arbitration' (1986) 2 Arbitration International 3,274 at
pp. 284-285; 0. Lando, 'Conflict-of-Law Rules for Arbitrators' in FestscJ~riftfiir Konrad Zweigert zum 70.
Geburtstag (ed. H. Bernstein et al.) (Tiibingen, Mohr 1981) at pp. 172-173.
However, it is far from clear how and to what extent an arbitrator must observe mandatory rules; see 0.
Chukwumerije, Choice of Law in International Commercial Arbih-ation (Westport, Greenwood Publishing
1994), at pp. 185-191.
If, in the case of an expressed choice of law by the parties, the arbitrators cannot leave open the question of
which conflict of laws rules they apply, then this must a fortiori be valid when there is only an implicit choice
of law.
This may not correspond with the usual practice in international trade. But nevertheless, it should not
necessarily be ascribed to the carelessness of the parties, or to the failure of the lawyers, since sometimes the
parties, after a long period of negotiation, cannot agree upon a choice of law clause. Instead of starting to
negotiate again they prefer to leave the issue open, or to postpone the problem to the time that a dispute
arises. rather than not to conclude the contract.
100
international lawT15 In other words, they could directly apply a substantive law
('voie directe').16A good example of such an approach can be found in section
1051(2) of the German Code of Civil Procedure (Zivilprozefiordnung)which
came into force on 1 January 1998." In this case, the problem of selecting a
conflict of laws rule would apparently not arise. Two possibilities are open to the
arbitrators when they apply a substantive law without any reference to conflict of
laws rules: the national law of a certain country, or a non-national set of rules (lex
mercatoria).
'" Seethe ICC Awards Nos. 1776/1966,3880/1982,4132/1983,4381/1986 cited by Berger, op. cit. atp. 499,
l6
l7
n. 128; Lew, op. cit. at p. 225; S. Toope, Mixed International Arbitration (Cambridge, CUP 1990) at p. 55.
See e.g. Art. 29(1) of the International Arbitration Rules of the American Arbitration Association: 'The
hibunal shall apply the substantive law or laws designated by the parties as applicable to the dispute. Failing
such a designation by the parties, the tribunal shall apply such law or laws as it determines appropriate'
(emphasis added). See also Art. 1486 of the French Code of Civil Procedure: 'The arbitrator shall decide the
dispute according to the rules of law chosen by the parties; in absence of such a choice, he shall decide
according to such rules as he deems appropriate'. See Art. 1054(2) of the Netherlands Arbitration Act.
This new provision reads as follows:
Section 1051 - Rules applicable to substance of dispute . . . (2) Failing any designation by the parties,
the arbitral tribunal shall apply the law of the State with which the subject-matter of the proceedings
is most closely connected.
Section 1051 is part of the amended Tenth Book of the German Code of Civil Procedure which contains
all German statutory provisions relating to aibitral issues. For a closer description and analysis of the new
German Arbitration Act see (1998) 14 Arbitration International 1 at p. 11. Although the wording of the new
section 1051(2) implies that arbitrators are entitled to apply substantive law directly to solve the dispute
concerned, it is far from clear whether the arbitrators are indeed released from having recourse to a conflict
of laws system, because the German legislature in its official reasoning proceeds on the assumption that
arbitrators have to apply the principles of German private international law. For a more detailed analysis of
the inconsistency of the wording of the new section with the legislature's intention, see D. Solomon, 'Das
vom Schiedsgericht in der Sache anzuwendende Recht nach dem Entwurf eines Gesetzes zur Neuregelung
des Schiedsverfahrensrechts'in (1997) RIW 981 et seq.
101
of law, namely the question with which mandatory provisions the arbitrators
have to comply. This difficulty can, as we have seen, only be solved by
applying some conflict of laws rules selecting the compulsory provisions.
Furthermore, if arbitrators apply the national law of a country without any
recourse to conflict of laws rules, they are always influenced by certain considerations having led them to exactly this particular system of law. These
considerations are nothing more than the application of conflict of Iaws rules
because the arbitrators will almost always look at factors, such as the 'closest
connection' of the contract, in order to determine the applicable law.'*
Therefore, the arbitrators de facto use conflict of laws considerations, even if
they are not aware of doing so. However, in this case it is better when they
unveil their deliberations not only to enable the parties to foresee the applicable law but also to make themselves realize that they are deciding a question
of conflict of laws. Hence, even if arbitrators choose a substantive law without
any reference to private international law, the selection of conflict of laws rules
is nevertheless necessary.
(ii) Lex mercatoria
Instead of selecting a national set of rules, the arbitrators may think of a nonnational standard.lg Although different labels such as transnational law,
international customs, lex mercatoria, and the like, have been used, they all
refer to the same phenomenon: a set of rules developed to regulate international
trade in the business community.20These substantive rules are derived not only
from international commercial dealings, standard clauses, international conventions and arbitral awards but also from various sets of legal rules issued by the
International Chamber of Commerce (ICC) or other international organi~ations.~'
l8
19
20
21
See ICC Award No. 423711984 in Jarvin and Derains, Collection of ICC Arbitral Awards 1974-1985
(Deventer, Kluwer Law International 1990) at pp. 170-171; ICC Award No. 6719/1991 in Arnaldez,
Derains and Hascher, Collection of ICCArbitral Awards 1991-1995 (Deventer, Kluwer Law International,
1997) at p. 577. Berger, op. cit. at p. 502: '. . . in international arbitral practice, the reasons which lead
arbitrators to select the appropriate applicable law are similar to the connecting factors used in conflict-oflaws rules'. Indeed, it seems as if the direct application of substantive law were akin to the English proper
law doctrine.
This choice may be excluded from the outset if the parties argue on the ground of national law, since this
may be construed as an exclusion of the application of non-national law. See Lando, 'Conflict-of-Law Rules'
op. cit. at p. 168.
B. Goldmann, 'The Applicable Law: General Principles of Law - The Lex Mercatoria' in Contemporary
Problems in International Arbitration (ed. J. Lew) (Dordrecht, Kluwer 1987) at pp. 113-116; C. Croff,
'The Applicable Law in an International Commercial Arbitration: Is It Still a Conflict of Laws Problem?'
(1982) Intern. Lawyer 613 at p. 623; M. Somarajah, International Commercial Arbitration: The Problem
of State Contracts (Singapore, 1990), pp. 116-117. The concept of lex mercatoria can in particular be
found in contracts between states or state entities and foreign companies because this supranational system
cannot be changed by the legislation of the state party. Hereinafter, the expression 'lex mercatoria' is used
synonymously for all labels.
See e.g. the ICC International Rules for the Interpretation of Trade Terms (INCOTERMS) and the
UNIDROIT Principles of International Commercial Conbacts; G.B. Born, International Commercial
Arbitration in the United States (Deventer/Boston, Kluwer Law International 1994) at p. 134.; H. van
102
Since only a state has the authority to enjoin a duty on a person, the
commitments resulting from a private contract are only derived from a state's
authority. The question of which country offers the framework for the principles
of the lex mercatoria must be decided on the basis of conflict of laws rules.
Another argument which has to be mentioned in this context is that an
arbitrator applying the lex mercatoria might find it impossible to decide every
aspect of the case by non-national standards because certain practices and
contractual clauses only apply in a specific industrial sector, or the principles are
frequently phrased in general terms, i.e. they are too elementary and platitudinous to permit detached evaluation of conflicting interest.24Although there
exist some common general principles, such as pacta sunt servanda, rebus sic
stantibus, the doctrine of the so-called 'effet utile9,etc., in practice it is difficult to
determine the 1ex mercatoria because its contents are vague and often
c~ntradictory.~~
When can pacta sunt servanda be set aside by rebus sic
stantibus? Some of the principles speak on such a high level of abstraction that
they afford little or no guidance in the resolution of concrete legal disputes. With
regard to international commercial contracts this existing vagueness might,
con t.
Houtte, 7he Law of International Trade (London, Sweet & Maxwell 1995) at para. 1.33; 'The UNIDROIT
Principles of International Commercial Contracts' (1995) 11 Arbiiration International 4 at 373; Berger, op.
cit. at pp. 526 et seq.; Michael J. Bonell, 'The UNIDROIT Principles of International Commercial
Contracts' in Process and Substance - Lectures on Comparative Law (ed. Roger Cotterrell) (London, 1995)
at pp. 45 et seq.
22 See Amin Rasheed Shipping Corp. v. Kuwait Insurance Co. [I9841 AC 50 at pp. 60 and 65, per Lord
Diplock; Redfern and Hunter, Law and Practice of International Commercial Arbitration (London, Sweet
& Maxwell 1991, 2nd ed.) at p. 71.
23
Saudi Arabia v. Aramco, decision of 23 August 1958, reprinted in (1963) 27 ILR 117 at p. 165.
24
see e.g. the standard conditions of the London Corn Trade Association, of the Grain and Feed Trade
Association (GAFTA), of the International Air Transport Association (IATA); van Houtte, op. cit. n. 2 at
para. 1.33; Redfern and Hunter, op. cit. at pp. 119-120. However, it must be admitted that gaps also
appear in national legal systems.
See ICC Award No. 423711984 in Jarvin and Derains, op. cit. at p. 170 rejecting the possibility of applying
the Iex mercatoria; Sornarajah, op. cit at p. 116 et seq.; van Houtte, op. cit. at paras. 1.33- 1.34 and 11.26;
Berger, op. cit. at pp. 544-553. See also the remark by Redfern and Hunter, op. cit. at p. 120: 'To this list
the cynical reader may feel like adding other well-known maxims of the "a stitch in time saves nine" variety!'
'"
103
26
27
28
29
30
31
The UNPDROIT Principles do not, for example, cover the transfer of property by sales and they exclude
issues of capacity and form; see van Houtte, (1995) 11 Arbitration International 4 373, at pp. 374 and 381.
Craig, Park and Paulsson, International Chamber of Commerce Arbitration (New York, London, Rome,
1990) at p. 295.
Berger, op. cit at p. 536; J. Hill, The Law Relating to In~ernationalCommercial Disputes (London, Uoyds
of London Press 1994) at p. 491; 0.Lando, 'The Lex Mercatoria in International Commercial Arbitration'
(1985) ICLQ 747 at pp. 764-768.
Tallon, quoted by J. Kropholler, Internationales Privatrecht (Tiibingen, Mohr 1990) at p. 396.
Mayer, op. cjt at pp. 284-296.
Lando, 'The Lex Mercatoria', op. cit. at p. 766; see also Bonell, op. cit at p. 71.
104
From the above it follows that an arbitrator always has to apply a conflict of laws
system. It does not matter at all whether the parties have explicitly or implicitly
chosen the applicable law, or whether the arbitrator directly applies a substantive
law without any recourse to private international law. Consequently, the next
logical step is to ask which conflict of laws system amongst those connected with
the dispute should arbitrators apply?34
32
33
34
35
The terms are used interchangeably. See e.g. Art. 13(4) of the ICC Arbitration Rules; Art. 28(3) of the
UNCITRAL Model Law; Art. 1054(3) of the Netherlands Arbitration Act; Art. 187(2) of the Swiss Private
International Law Statute; 5 21.3 of the Arbitration Rules of the German Institute for Arbitration.
Berger, op. cit at p. 574; Hill, op. cit at p. 487; Craig, Park and Paulsson, op. cit. at pp. 313-314.
his problem will not occur if the parties agree upon the applicable private international law; see e.g. ICC
Award No. 211411972 in Jarvin and Derains, op. cit at p. 52.
Art. 13(3) of the ICC Arbitration Rules which was influenced by Art. VII(1) of the European Convention
on International Commercial Arbitration; Art. 33(1) of the UNCITRAL Arbitration Rules; Art. 33(1) of the
Inter-American Commercial Arbitration Association; Art. 46 of the Netherlands Arbitration Institute
Arbitration Rules; 5 21.2 of the Arbitration Rules of the German Institute for Arbitration; see 0. Sandrock,
Welches Kollisionsrecht hat ein Internationales Schiedsgericht anzuwenden?' (1992) RIW 785 at pp. 790791 with many further examples. See on the other hand, Art 42(1) of the Convention on the Settlement of
Investment Disputes between States and Nationals of Other States, which provides that the tribunal must
apply the conflict rules of the contracting state which is a party to the dispute.
105
(a) Law o f the Country which would have had Jurisdiction in the Absence o f a n
Arbitration Clause
One of the first scholars who dealt with the problem was Dionisio A n ~ i l o t t iHe
.~~
held the opinion that arbitration should be governed by the conflict of laws
system of the country which would have had jurisdiction to settle the dispute
between the parties, had they not included the arbitration clause in their
contract, for the arbitration agreement has dispossessed the jurisdictional
authority of that country and as a consequence it may reaffirm its control over
arbitration in this way.3g
This theory can be criticized on several grounds. First of all, there arises the
problem of which country would have had jurisdiction over the dispute. In
practice, this question is not as easy to answer as it might appear reading
Anzilotti's statement, for in international trade more than one country might
have jurisdiction over the dispute. Thus, it would be up the arbitrator to decide
which country is the relevant one in the argument. One major advantage of
arbitration - the avoidance of uncertainty and inconvenience relating to
international conflicts of jurisdiction - would be totally f r ~ s t r a t e dTherefore,
.~~
the unpredictability of the dispute's result would increase. Finally, this solution is
not acceptable because it is circular. In order to ascertain the country which, in
the absence of the arbitration agreement, would have had jurisdiction, the
arbitrator also has to select a kind of conflict of laws rule. For this reason, the same
problem would arise for which the theory strives to provide a solution.
36
37
38
39
"
British Petroleum v. Libya (Merits), decisions of 10 October 1973 and 1 August 1974, reprinted in (1979)
53 ILR 297, at p. 326.
Jaffey, op, cit. at p. 132; Lando, 'Conflict-of-Law Rules', op. cit. at p. 159.
(1906) Rivista di Diritto Internazionale 467.
See also in a different context the formulation by Hill, op. cit. at p. 492: '. .. in international situations the
arbitrator takes the place of all the courts which might have had jurisdiction to determine the dispute in
the absence of the arbitration agreement.'
Oroff, op. cit. at p. 624.
106
Consequently, this position has never formed the basis of any known arbitral
awardm4'
(b) Seat o f Arbitration
The traditional point of view is that the conflict of laws rules should be applied
which pertain to the country in which the arbitral tribunal is located." The basis
for this opinion is the assumption that a national court and an arbitrator play the
same role in a legal system, i.e. the arbitrator replaces the judge of a national
court. Accordingly, it is a small step to conclude that if the judge has to apply the
private international law of the forum, the same must mutatis mutandis be right
for the arbitrator. Therefore, he has to make use of the private international law
of his 'natural forum', the lex arbitri." Another argument in favour of this
approach is the correspondence of the procedural law of arbitration (the lex
arbitr~)with the applicable conflict rule.
This point of view does not only entail advantages but can also be challenged
upon both practical and theoretical grounds. First of all, it must be mentioned
that this solution furthers one of the major advantages of arbitration because a
strict application of this 'seat theory' entails the predictability and uniformity of
disputes' results. Furthermore, the will of the parties is respected: by choosing
the seat of arbitration, they can indirectly select the applicable conflict of laws
rules. The last point, however, also implies disadvantages, because if the parties
have failed to agree upon the seat of arbitration it is up to either the arbitrator44
or an arbitral institution such as the ICC" to choose it instead. In this case, the
argument of respecting the parties' wishes does not carry weight. Moreover,
the conflict of laws rules of the tribunal's seat will normally have no connection at
all with the dispute in question. Worse, applying the conflict of laws rules of the
seat might result in the application of a law unintended by either party. This can
be demonstrated by an example." An arbitral tribunal domiciled in Germany
has to decide upon a dispute between a company from Syria and a company
from the USA. On the failure of the parties to choose the applicable substantive
law, the applicable law is determined under the Syrian conflict rules" by the lex
loci contractus. Under the conflict of laws rules in the USA, the place of
4'
43
44
45
''
ibid.
ICC Award No. 546011987 in Jarvin, Derains and Arnaldez, Collection of ICC Arbitral Awards 1986-1990
(Deventer, Kluwer Law International 1994) at p. 138.
The leading exponent of this traditional view was F.A. Mann, 'Lex Facit Arbitrum' in International
Arbitration: Liber Amicorum for Martin Dornke (ed. P. Sanders) (The Hague, Kluwer 1967) at p. 167; for
ICC cases applying the 'seat theory' see Lew, op. cit at pp. 201-202, 239-240 and 255-272. See also
Czar~kowv.Roth, Schmidt & Co. 119221 2 KB 478 at p. 488, per Scrutton LJ.
See e.g. Art. 20(1) of the UNCITRAL Model Law on International Commercial Arbitration.
Art. 12 of the ICC Arbitration Rules; see also Art. 7.1 of the Rules of the London Court of International
Arbitration (LCIA).
See Sandrock, op. cit at p. 788; see also the example by EJ. Cohn, 'The Rules of Arbitration of the
International Chamber of Commerce' (1965) ICLQ 132 at p. 162.
Art. 20(1) of the Syrian Civil Code 1949.
107
contracting also gains some importance to identify the applicable law." Perhaps
the parties signed the contract in either Damascus or Detroit, being conscious of
the significance of the place of contracting for the determination of the
applicable law. In contrast to this, the German conflict o l laws rulea as the
applicable provision in accordance with the 'seat theory' does not emphasize the
place of signature in order to find the substantive law of the ~ontract.~"
Consequently, applying German conflict of laws rules would result in completely
opposing the parties' intentions.
In addition, if the arbitrator does not expressly lay down the seat of
arbitration, its identification may be impossible. For practical reasons the
arbitrator might have fixed the hearings in two or even more different countries
because the evidence pertaining to each country is more accessible that way. It is
not a convincing solution to this problem to say that the law of the country where
the first meeting was held shall prevail, since this is merely a static and arbitrary
selection. If the arbitration is solely conducted by an exchange of letters between
the arbitrator and the parties, the problem becomes even more difficult. Where
does the arbitration take place in this case? At the domicile or the residence of
the arbitrator, or of one of the parties? Neither selection would be satisfactory.
Indeed, if there are sundry possibilities open for the seat of arbitration the
solution of this theory is also circular because in order to select the appropriate
seat of the arbitral tribunal a kind of conflict of laws rule is again necessary.
Then, the theory would fail in exactly that case for which it strives to give an
answer.
If an institution determines the place of arbitration the former problems are
less likely to occur because it will clearly state the seat of arbitration. Then, it
does not matter when the hearings are held at different places. However, in this
case, the advantages of certainty inherent in the 'seat theory' are less obvious,
since, until the institution has laid down the arbitral tribunal's seat, the parties
will not know the place, and thereby they cannot foresee the substantive law
applicable to the merits of their dispute. This period of time is therefore wasted
for them in order to prepare the case.51 Hence, the main advantage of this
position, i.e. the predictability of the applicable substantive law and by that the
potential result of the dispute, only prevails as long as the parties themselves have
chosen the seat of arbitration.
Another aspect of this theory causing criticism is based on practical grounds.
The reasons why the parties or the arbitrator choose a certain country as the seat
of arbitration usually have nothing in common with the conflict of laws provisions of that country. Therefore, the rule qui eligit arbitrum eligit ius basically
" Restatement of the Law 2d, Conflict of Laws, Vol. 1, by the American Law Institute, s. 188(2)(a) and (3).
4g
50
51
Art. 28 of the In&oductory Law of the Civil Code is the same as Art. 4 of the Rome Convention.
Art. 4(2) of the Rome Convention.
See K.H. BGckstiegel, Arbitration and State Enterprises (Deventer, Kluwer Law International 1984) at
p. 27; Craig, Park and Paulsson, op. cit at p. 136.
108
does not apply to international arbitration, and a fortiori does not apply to the
determination of the appropriate conflict of laws rule, although the choice of the
seat of the arbitral tribunal may be one factor among others taken into account
when looking for an implicit choice of law.52The parties might even not be aware
that a selection of the place of arbitration can become so important for the
solution of their possible future disputes. On the contrary, the seat of arbitration
might be chosen because its location is for all parties most con-venient. It is, for
example, for all participants of the arbitration the closest place; it provides the
necessary infrastructure, such as conference rooms, interpreters, etc.; or it is a
neutral place, i.e. neither party has any connection with that place or could take
advantage of 'playing at home'.53
Last but not least, there is a more theoretical reason demonstrating a weakness
of the 'seat theory'. When it is maintained that arbitral tribunals are comparable
to national courts then the argument that arbitration replaces litigation is overemphasized, since both derive their powers from different sources: the private
tribunal derives its power from the arbitration agreement, whereas a national
court derives its power from the state and thereby from the public. In contrast to
this, arbitrators do not exercise public or institutional powers in the name of a
state. As a consequence, the basis of the 'seat theory', namely the comparison
with national judges, neglects the private character of arbitration. Summing up, it
has to be said that the advantages of the seat approach, namely the implied
predictability of disputes' results, only carry weight as long as the parties have
agreed upon the seat of arbitration. Otherwise, it appears that the disadvantages
of the theory prevail.
53
54
55
Cmpagnie D'Armement Maritime S.A. v. Compagnie Tunisienne de Navigation S.A. [I9711 AC 572 at
p. 588, per Lord Morris and at p. 597, per Lord Wiberforce. Dicey & Morris, op. cit. at p. 761; U.S.
Supreme Court, Scherk v. Alberto Culver, 417 US 506 at 519 n. 13 (1974). In contrast to this, see the
German Federal Supreme Court (Bundesgerichtshofl (1967) AWD/RIW 108 at p. 109: 'In agreeing on an
arbitral institution, the parties also agree on the application of the substantive law in force at its seat' (original
in German). But looking at the decision as a whole, it becomes obvious that in fact the court only used the
seat of arbibation as one indication of the parties' will; see Berger, op. cit. at p. 496 n. 108.
See Lando, 'The Law Applicable to the Merits of the Dispute', op. cit. at p. 102;van Houtte, op. cit at para
11.22, 'equality of arms', and at para. 11.23 for legal considerations concerning a suitable law of arbitration.
see Craig, Park and Paulsson, op. cit. at p. 288, n. 14.
See Croff, op. cit. at p. 628.
109
proposed rule would be unsuccessful in just that case for which it attempts to find
a resolution. These problems would increase if an arbitrator is replaced by
another of different nationality, or if the tribunal consists of three arbitrators.
Should, then, the nationality of the majority of arbitrators be decisive, or that of
the chairman? Moreover, one would not do justice to international arbitrators if
their capacity to understand and correctly apply foreign private international law
were underestimated. Conversely, arbitrators in international commercial arbitration are almost always appointed because of their experiences with different
legal systems.56
Similarly, another approach has to be rejected: the arbitrator cannot simply
apply the private international law of one of the parties because, regularly coming
from different countries, at least one party is left unsatisfied. A further attempt to
use the conflict of laws system of the country where the award will be enforced is
also not convincing. Though the arbitrator shall make every effort under Article
26 of the ICC Arbitration Rules to sender an enforceable award, which he is
more likely to achieve if he applies the private international law of the country of
enforcement, he will hardly foresee the forum of enforcement because the losing
party may have assets in different countries. Since the theories which favour a
single conflict of laws system are not entirely convincing, one has to turn to
approaches concentrating on a broader basis.
5"ee
57
110
decision on all conflict of laws systems related to the dispute. But it cannot be
denied that the application of the theory is confined to situations of 'false
conflict'. Therefore, this position does not provide a solution when the private
international laws of the countries connected with the dispute lead to different
substantive laws. Furthermore, an arbitrator will have discretion in order to
establish which conflict of laws systems he has to take into account. Thus, for
example, a person who stresses the contractual nature of arbitration might
neglect the conflict of laws rules of the lex a r b i t ~ iIn
. ~ addition
~
to that, the likely
long-lasting review of these legal orders will regularly delay the process of arbitration and make it more complicated, which contrasts sharply with one of the
parties' motives in choosing a r b i t r a t i ~ n . ~ ~
A similar approach is the application of an international conflict of laws
system allowing the arbitrator personally to create a conflict of laws rule with an
international ~haracter.~'
A significant example is the ICC Award No. 46501
1985:62'The arbitral tribunal does not deem it necessary in this case to decide on
a specific rule of conflict to designate the proper law of the contract in view of
the fact that most major rules in some form or another point to the place of the
characteristic or dominant work . . .'. The tribunals then apply those conflict of
laws rules which are common to all or at least the leading legal systems and to
international convention^.^^ Both this theory and the approach which applies
conflict of laws systems cumulatively are based on a comparative analysis of
several bodies of private international law and thereby simultaneously on the
conflict of laws rules of the countries connected with the dispute. The difference
is that the arbitrator applying a cumulative approach can leave the answer open
as to which systems he finally applies, whereas the arbitrator applying an international system must find a compromise among different laws. Since in the latter
case the arbitrator has to decide which general principle of international private
law prevails, the latter position also applies if there is no 'false conflict'.
A further approach encouraging 'denationalized' conflict of laws rules is that
arbitrators should apply conflict of laws rules which are supposedly contained in
public international law, in order to determine the applicable law.64 This
viewpoint can be challenged on practical grounds because attempts to discover
conflict of laws rules laid down by public international law have not been
59
63
64
See ICC Cases Nos. 1759 and 1990 of 1972 cited by Craig, Park and Paulsson, op. cit at p. 291, n. 24.
Born, op. cit. at p. 115; Berger, op. cit at p. 505.
ICC Cases Nos. 209611972 and 2585/1977 cited by Craig, Park and Paulsson, op. cit. at p. 291, nn. 26 and
27. See Jaffey, op. cit at p. 145; Lew, op. cit. at pp. 327-335.
In Jarvin, Derains and Arnaldez, op. cit. at p. 68.
See the ICC Awards in Jarvin and Derains, op. cit., No. 171711972, p. 191; No. 268011977, p. 30% and
No. 331611979, p. 88; see also the ICC Awards in Jarvin, Derains and Arnaldez, op. cit. No. 571311989,
p. 223; and No. 628111989, p. 249; Lew, op. cit. at p. 584.
See e.g. Professor Dupuy's statement in Texaco Overseas Petroleum Co. v. Libya (Merits), decision of 19
January 1977, reprinted in (1979) 52 ILR 422 at p. 445; Libyan American Oil Co. v. The Government of
the Libyan Arab Republic (Merits), decision of 12 April 1977, reprinted in (1981) 20 ILM 42, although
there it was not deemed necessary to apply such rules.
111
successful. Since public international law cannot provide real guidance, purporting to choose the conflict of laws rules of international law is, in reality,
nothing more than a veiled attempt to allow arbitrators to apply any substantive
law they wish." Therefore, arbitrators nevertheless have to choose certain
conflict of laws rules, at least to determine the mandatory rules which are
decisive for the dispute in question, because in this case the same principle is
relevant as if the arbitrators had applied the substantive law without any recourse
to private international laweG6
IV.CONCLUSION
Several conclusions can be drawn from the above. First, an arbitrator cannot
ignore the question of private international law in order to determine the
substantive law applicable to the merits of the dispute. In other words, he cannot
simply rely on the choice of law by the parties, or apply a national or nonnational system of law which he regards as appropriate for the solution of the
dispute. Even then, the question which conflict of laws rules apply has to be
answered. Secondly, under the auspices of almost all leading institutional
arbitration rules the arbitrator can apply the rule which he deems appropriate.
The (far too) wide discretion offered by these rules may lead to the application of
an unforeseeable set of conflict of laws rules. In order to preserve the predictability of an arbitral decision one must therefore look for an additional stage in
the hierarchy, i.e. a kind of ultimate conflict of laws rule, leading to the appropriate one, which then determines the substantive law applicable to the merits of
the dispute.
From this, one can finally gather that no theory alone provides a satisfactory
solution for this problem. All attempts comprise both advantages and disadvantages, even though the latter prevail in some theories, such as the
approach by Anzilotti, or those which regard the law of the arbitrator or the law
of the place of enforcement as decisive. As we have seen, the advantages of
both the 'seat theory' and the cumulative approach can only be acknowledged
to a certain extent. Therefore it may be maintained that the best approach is to
combine these different views because it may result in avoiding or compensating the disadvantages involved in each theory alone. Then the disadvantages
of these views either can be ignored because they are of minor importance, or
are balanced out because of the combined application of the theories. The
cumulative approach entails the problem that it only applies if there is a real
'false conflict' among the conflict of laws rules connected with the dispute. In
contrast to that, the disadvantage that arbitrators might not select the private
international laws in a uniform way can be discounted because the scope of
" Toope, op. cit at p.
'"ee
51.
112
113
dl, the arbitrator may simply count the laws leading to the same substantive law
and then apply the principles of this majority. Secondly, one can imagine
applying the rules of that country having the closest connection with the dispute.
Finally, it is imaginable that a certain convention will be applied, such as the
Rome Convention. Ultimately, this is a step towards an internalional conflict of
laws system. However, by applying only the rule of a national legal system or of a
convention, the arbitrator will not operate in a legal vacuum. This last stage of an
ultimate conflict of laws rule leading to the appropriate rule is thus not a 'new
sort of lex mercatoria' for private international law independent from all legal
systems. The measure of unpredictability possibly entailing this solution may be
ignored in order to find a suitable system of conflict of laws rules. Besides, the
number of cases in which the last stage of this combined approach has to be
taken may be of less importance because a solution will almost always be found
within the first two steps.
In summing up, it may be said that the best solution to determine the substantive law applicable to the merits of the dispute is to combine three of the
approaches offered in order to eliminate their disadvantages and to further their
advantages. This solution itself might provide flexibility while at the same time
ensuring that the arbitrators are not left without guiding principles. Hence, it may
also lead to a compromise between two conflicting positions: on the one hand,
the wish to achieve a perfect and stringent system of rules, i.e. a maximum of
certainty and predictability; on the other, the quest for sufficient creativeness and
inventiveness, leaving enough room for the arbitrators to take the facts, circumstances and interests involved in each individual case into consideration.
This combination is an ultimate conflict of laws rule, leading to the appropriate
rules which determine the law applicable to the merits of the dispute. The result
is a three-stage procedure:
(2) application of the rules of conflict of laws being in effect at the seat of
arbitration if the parties have chosen it; if not
(3) application of the general principles of private international law which can
be found by a comparative review of the conflict of laws rules connected
with the dispute.