Edited by Jürgen Basedow, Giesela Rühl, Franco Ferrari and Pedro de Miguel Asensio
Chapter F.5: Floating choice of law
I. Introduction and cases of disputed contractual consent
There is a strand woven through the private international law of contract which is unproblematic all the while only glancing attention is paid to it but which, when subjected to closer scrutiny, reveals problems of flaws which run deep in the methodology of the subject. For example, the procedure by which the law determines whether the parties made a contract is generally said to be answered by recourse to the law which would have governed the contract if it were valid, and using this law to determine whether the contract was valid. At first sight it seems entirely sensible that the law which governs the proposed contract also be the law to determine whether there was a valid and binding contract in the first place. But the lack of logic, and the prejudice, inherent in this, does not take long to reveal itself. It lacks logic, for there is no obvious reason to approach the question on a footing which asks how the issues would have been resolved if there were a contract before it has been decided that there was a contract in the first place: it seems unsatisfactory to say, in effect, that the court should assume that there was a contract in order for it to decide whether there was a contract in the first place. And it involves prejudice, for it involves making some kind of assumption which is aligned to the contentions of the party who asserts that there was p. 766a contract, and which is opposed to those of the party who denies it.
Yet if the response to this is that one should then start the inquiry by making reference to the law of the forum, as a point of material contact which does not depend on whether there was a contract at all, the risk is that the result of commercial litigation will be dependent on the forum in which the proceedings are brought, which is also unsatisfactory. No national system for →choice of law was altogether convincing in the manner in which it dealt with this threshold question. Of course, if the parties to litigation agree that there was a contract, the problem does not arise, but if they take issue on this fundamental, threshold, point, the issue can appear to be very difficult indeed. For these among other reasons, the solution which was adopted in the Rome Convention (Rome Convention on the law applicable to contractual obligations (consolidated version),  OJ C 27/34), and reaffirmed in the Rome I Regulation (Regulation (EC) No 593/2008 of the European Parliament and of the Council of 17 June 2008 on the law applicable to contractual obligations (Rome I),  OJ L 177/6; →Rome Convention and Rome I Regulation (contractual obligations)), was to subject the issue of contractual formation to a statutory, two-stage test: to apply the law which would have governed the contract if it had been assumed to be valid, but to allow a party whose own law would have reassured him that he had not consented to be bound to rely on this law in circumstances in which it is reasonable for him to do so (art 8 Rome Convention; art 10 Rome I Regulation). Like most two-stage tests, it lacks a certain degree of elegance, but it tends to result in answers which are intuitively correct; and it is undeniable that the imposition of a choice-of-law rule from an external source avoids the problem of reasoning which pulls itself up by its bootstraps.
II. Floating choice of law and uncertainty in national law
With that illustrative introduction we may proceed to examine the issue of the so-called ‘floating’ choice of law. Situations which fall within this descriptive category will include those in which (i) a contract contains a provision according to which the contract is to have no initial governing law, but provides that the governing law may be specified (whether selected from a limited menu of possibilities or otherwise) at a later date, or (ii) a contract provides that it will be governed by one law unless and until a party specifies that it is to be governed by another law, or (iii) a contract provides that it shall be governed by the law of country A or country B, at the election of one of the parties, or (iv) a contract provides that its governing law shall be that of the place (whether this be selected from a defined menu of specific possibilities or otherwise) where dispute resolution is to take place. What these several variants have in common is that the choice by the parties of the law to govern the contract cannot be said to be clear, fixed and determined from the outset, but is by their express choice left open to be determined later, or made provisional and open to being altered or departed from later. It is also important to realize that such agreements are rare, and that when they are made this is unlikely to have come about by inadvertence. Whatever is said about them, therefore, is only of marginal significance in the development of private international law.
A view of first impression might be that if this ‘floating choice of law’ is what the contracting parties wanted, there needs to be a compelling reason, doctrinal or practical, why they should not be allowed to agree to it and have it applied to their contract. But the meagre authority available in the common law, at least, took (and in the countries in which it prevails, may still take) the opposite view. If it were not permitted for there to be a ‘contract without a law’, it would follow that it was also impossible, as a matter of legal doctrine, for there to be a contract temporarily without a law, or a contract which, for a period of time, would have no governing law. If, therefore, the parties had agreed that governing law would only be capable of ascertainment at some point, upon the occurrence of some event, after the making of the contract in the first place, they had purported to create something which it was beyond their legal power to manufacture. The conclusion, which might have been thought to follow, was that the entire agreement was a nullity: either on the basis that if the provision which was to be its lynchpin was omitted, there was nothing to hold any of the rest of it together, or because the conclusion that there was a contract but one governed by a law which, ex hypothesi, was contrary to their choice, was not legally coherent. But a perception that this would be taking the logic of doctrine rather too far led to its being p. 767generally accepted that the contract would take effect minus the ineffective provision for a floating choice of law. The contract would be governed as a result by the law with which the contract had, or otherwise would have had, its closest and most real connection, that law being assessed by reference to the facts and matters as they stood at the date of the making of the contract.
By contrast with this approach and conclusion said to be derived from it, it appears to have been rather more easily accepted in civilian systems of states party to the Rome Convention that a post-contractual →choice of law, or variation of earlier choice of law, was permissible. The laws of →Germany and the →Netherlands, as well as of →Switzerland, appear to have been prepared to allow a post-formation choice of law, and to conclude that the law thus and then chosen would apply, with retrospective effect, to govern the contract, including questions concerning its formation. However, Italian law was not so obviously aligned with this approach, and in truth, there was rather little authority of any kind, in any system, to vouch for the correct result.
Among the more curious aspects of the view that a floating choice of law is not to be permitted is its contrast with the private international law of →torts. For it was never said that there was any great difficulty with the fact that the relationship between the parties to a tort might not yield or disclose its governing law until the parties reached the point of litigation. After all, some torts arise from a prior relationship of some duration, just as contracts may. If it was not seen as a problem that a potential tortfeasor would not know until it was too late to do anything about it which law would be applied by a court to determine, with retrospective effect, whether he had complied with the particular requirements of care or safety, or which damaging or loss-causing consequences he might be held responsible for, it is not obvious why matters should stand any differently if the relationship is a contractual one instead. Again, it is sufficient that the law be ascertained when the parties come to litigation; and if that is so for torts, why not for contracts as well?
An answer to that question, sometimes advanced, is that contracts, and contractual relationships, are different, and that parties to contracts need to know during their performance of the obligations which they have undertaken which law will provide the background against which these obligations are to be interpreted. This, however, is not completely persuasive. If the parties have spelled out the terms of their agreement in such a way as will mean that the identification of a governing law is not as easy as it might be, then if either of them is inconvenienced he has only himself to blame. It is much more likely that the specification of a choice of law as floating, to be determined later, is done for good commercial reasons. These may include making it simpler to litigate, by allowing the applicable law to follow the place of litigation, or by allowing parties, who apprehend that a law which they considered choosing might be changed, to insulate their contract from the risk of unwelcome legislative developments. For example, it might be that the parties wish to stipulate for a number of permitted fora for litigation, but to provide also that the law which applies shall be the law of the place where litigation takes place. The interest of the law in denying the parties the opportunity to organize their dispute-resolution on this basis is, surely, low.
A second curious aspect of the proposition that a floating choice of contractual law is not permitted lies in the contrast this poses with the fact that the law governing an arbitration, or the curial law of the arbitration, may be left floating until such time as the claimant decides to institute arbitral proceedings before whichever tribunal is opened to him. It is usual that the curial law be the law of the place where the arbitration is to take place; it has never been seriously contended that there is any problem with the proposition that the curial law may, in this sense, be floating until the arbitration is instituted.
If one regards the law which governs a contract as an indispensable part of a contractual relationship, which cannot be formed without one, then denying the possibility of a floating choice of law would make some sense. But in most cases, the truth is that the law which governs the contract is of practical relevance only at the point of the resolution of disputes, and that if the identification of that law is left until the process of dispute resolution has been opened, no great or obvious harm is done. After all, it is perfectly possible to see the choice of law by the parties to a contract as a choice of the law which they wish to have applied by the court or tribunal which deals with disputes between them.
However, if it is accepted that a contract must have a law from the beginning, and the parties have provided for a law to be chosen at a later p. 768date, the next question must presumably be to ask whether the law which governs the contract from the beginning permits (or imposes limits upon the freedom of) the parties to exercise a choice and so to change the governing law. At this point it is as well to acknowledge that logic is liable to lead the law over a cliff, and that the cause of the problem, as shown at the beginning of this section is that if the parties’ autonomy (→Party autonomy) is to be the source of their power to make a choice, there has to be law by which to measure the nature and extent of their autonomy. Autonomy, as Professor Nygh pointed out, cannot provide its own justification, any more than a man can pull himself up by his own bootstraps.
III. Legislative solution
An external, legislated, rule for →choice of law is required to provide authority for the choice of law, and for such variations in the mode and manner and timing of choice of law which the parties may make. For the Member States of the European Union, this is now provided in the welcome form of the Rome I Regulation. If the parties express a choice of law at the outset, in a manner which conforms to art 3, well and good. If they do not, the law which governs the contract will be that identified by art 4 of the Regulation; but as art 3(2) allows the parties to change the law which governs their contract at any time, and (so long as they choose the law of a state) for any law, a post-formation choice of law will be effective because the legislative scheme of choice of law confirms that it is valid without further ado.
(1) A contract shall be governed by the law chosen by the parties. The choice shall be made expressly or clearly demonstrated by the terms of the contract or the circumstances of the case. By their choice the parties can select the law applicable to the whole or to part only of the contract. (2) The parties may at any time agree to subject the contract to a law other than that which previously governed it, whether as a result of an earlier choice made under this Article or of other provisions of this Regulation. Any change in the law to be applied that is made after the conclusion of the contract shall not prejudice its formal validity under art 11 or adversely affect the rights of third parties.
At first sight, this means that the parties to the contract may adopt a common position that, no matter what the law governing the contract may have been, and no matter what it might say about the freedom of the parties to choose another law, the lex contractus will henceforth be a different law. They do not need to derive authority for this exercise of choice from the law which previously governed the contract, for the right to agree is conferred upon them ab extra, by the legislative wording of the Regulation itself. But if its wording is taken at face value, art 3(2) Rome I Regulation can really only apply in a case in which the parties join in a common act of variation: its wording does not appear to be apt to cover a case in which one party exercises or purports to exercise a right to select the governing law. Where, therefore, the contract contained its own mechanism for making a choice or alteration of law, giving that option or power to one party, art 3(2) is not obviously applicable, and the justification for or validation of the new stipulation of governing law must be found in the original contractual power. For making a choice or change of governing law pursuant to a contractual power is different and distinct from what happens when a lex contractus is altered by the joint act of the parties to the contract.
In the final analysis, the arguments in favour of allowing the parties to postpone the identification of the law which will govern their contract, and to provide a default solution to fill any hiatus in governing law, far outweigh any which oppose this on doctrinal grounds.
Michael N Howard, ‘Floating Choice of Law Clauses’  LMCLQ 1;
Otto Kahn-Freund, ‘General Problems of Private International Law’ (1974-III) 143 Rec. des Cours 139;
Otto Kahn-Freund, General Problems of Private International Law (Sijthoff Leiden 1976);
Ole Lando, ‘Contracts’ in Konrad Zweigert and Ulrich Drobnich (eds), International Encyclopaedia of Comparative Law, vol III ch 24 (Mohr Siebeck 1976) 51;
Francis A Mann ‘The Time Element in the Conflict of Laws’ (1954) 31 BYIL 222;
Peter North, Essays in Private International Law (OUP 1993) ch 3;
Peter Nygh, Autonomy in International Contracts (OUP 1999) esp 98–100;
Richard Plender and Michael Wilderspin, The European Private International Law of Obligations (3rd edn, Sweet & Maxwell 2009);
Józef Skapski, ‘Les obligations en droit international privé dans le cadre du commerce extérieur des pays socialistes’ (1972-II) 136 Rec. des Cours 499;
Maciej Tomaszewski, ‘Le désignation postérieur à la conclusion du contrat de la loi qui le régit’  Rev.crit.DIP 567.