The Rome I Regulation (593/2008/EC) is the EU regulation which governs the choice of law applicable to contractual obligations in civil and commercial matters. It replaces the 1980 Rome Convention and effects important structural changes to the rules for determining the applicable law, as well as the exceptions to those rules (the ‘choice of law gateways’). The English courts’ application of the gateways in the Rome Convention has been criticised as excessively flexible. While this view may be justified in relation to earlier cases decided on the Convention, there has been a gradual movement towards a more cautious judicial approach to both implied choice (Article 3) and the escape clause (Article 4), which was partly crystallised and partly precipitated by the changed architecture of these two provisions in the Rome I Regulation.
I. Implied choice – Article 3 Rome I
Article 3(1) Rome I stipulates that where the parties have not made an express choice of law, courts may imply such a choice if it is ‘clearly demonstrated by the terms of the contract or the circumstances of the case’. This is a notable change from Article 3 of the Convention which only required that a real choice be ‘demonstrated with reasonable certainty’.
(a) 'Terms of the contract'
The most compelling evidence of a real but unexpress choice is the inclusion of a jurisdiction or arbitration clause as a contractual term. Jonathan Hill notes that in such cases, English courts imply a choice of the law of the forum ‘almost as a matter of course’.1 Yet judges are arguably not abusing their discretion in doing so because (a) such an approach is sanctioned by Recital 12 of Rome I, and (b) other EU Member States do the same (notably Germany). Furthermore, an exclusive jurisdiction clause is arguably most indicative – out of all the factors which a court may consider – that the parties intended to make a choice of law, namely the law of the chosen forum. Therefore, in this respect, English judges’ use of the implied choice gateway is consistent with Article 3 and the overarching principle of the Rome regime: party autonomy.
Nevertheless, English courts’ propensity to find implied choice of law on the basis of factors other than choice of forum clauses has sometimes been unwarranted. A notable example is the House of Lords’ judgment in Amin Rasheed Shipping Corp v Kuwait Insurance Co  3 WLR 241. Admittedly, Lord Diplock’s reasoning that contracts cannot exist in a ‘legal vacuum’ was founded on impeccable legal logic. However, the House of Lords’ decision to imply a choice of English law because the contract was based on the Lloyds marine policy standard form – which had to be interpreted by reference to the Marine Insurance Act 1906 and the English case law on it – was a stretch too far. Lord Diplock noted (quoting Bingham J, as he then was, in an earlier case) that the Lloyds policy was so widely used that it had become a ‘common currency’ for the international insurance business. Contrary to his view, however, precisely because it is so widely used it cannot realistically be presumed that all private parties in the world who choose the Lloyds form as their insurance policy also intend to choose English law to apply to their contractual disputes, if nothing else in the terms of the contract or the circumstances of the case indicates such a choice.
(b) 'Circumstances of the case'
When applying this second discretionary limb of the implied choice gateway, English courts have shown more restraint. While Potter LJ in Aeolian Shipping SA v ISS Machinery Services Ltd  EWCA Civ 1162 reasoned that a wider range of factors could be considered under Article 3 than under the traditional common law approach to implication of terms, Mance LJ in American Motorists Insurance Co v Cellstar Corporation  EWCA Civ 206 disagreed and formulated the ‘so obvious that it went without saying’ test. Its application in subsequent cases allowed English judges to remain within the boundaries of implication of a real choice, and not stray into the proscribed waters of imputation of a choice which was clearly not made but would have been reasonable. In light of this, the English courts’ approach to Article 3 has been overall satisfactory and there has not been overindulgence in search for an implied choice. This relatively strict approach of the English judges was crystallised by the new architecture of Rome I, whose emphasis on the objectives of certainty and predictability necessitates the curbing of interpretative flexibility. Therefore, it is unlikely that there will be a major change in the English courts’ current application of the implied choice gateway.
II. ‘manifestly more closely connected law’ – Article 4 Rome I
In contrast to their measured approach to Article 3, English courts have adopted too malleable an interpretation of Article 4, which applies when the parties have not made a choice of law. This is reflected in the higher threshold for the activation of the choice of law gateway in Rome I. First, the presumption in Article 4(2) of the Rome Convention that the law of the country where the characteristic performer habitually resides shall govern the contract has been raised to a rule in Rome I. And second, the escape clause in Article 4(5) of the Rome Convention, which enabled the displacement of the presumption in favour of another more closely connected law, has been reduced to a narrow exception in Article 4(3) Rome I.
In the early Article 4 cases, English courts tended to find that the place of performance of the obligations characterising the contract was a more important localising factor than the place of habitual residence of the party performing it, and to conclude that the contract was more closely connected with the law of the former. In Samcrete Egypt Engineers & Contractors SAE v Land Rover Exports Ltd  EWCA Civ 2019, the Court of Appeal found that English law rather than Egyptian law (where the characteristic performer had their habitual residence) governed the contract because the place of delivery and payment was England. Yet apart from these two connecting elements, there was nothing else linking the disputed contract to England. Indeed, as Hill argues, the court attached so little weight to the country of habitual residence of the characteristic performer that it can almost be said that it was presuming – contrary to the letter of the Convention – that the place of performance was the best indicator of the closest connection between law and contract.2
The judgment of Morison J in Definitely Maybe (Touring) Ltd v Marek Lieberberg Konzertagentur GmbH  1 WLR 1745 was an attempt to mediate between the weak and strong versions of the presumption. His strictly textual approach to construction led him to hold that Article 4(2) was intended to operate as a ‘normal rule’ which would be ‘simple to apply’ and would apply in the majority of cases, thereby bringing the English approach to Article 4 closer to the strong theory. Nevertheless, in judging that the presumption could be disregarded in cases where the factors connecting the contract to another country overwhelmingly prevail, he went on to do just that (interestingly, with the effect of stripping English courts of jurisdiction). In light of this judgment, and in line with Hill’s view, it appears that while English judges are speaking the language of the strong version of the presumption, in practice they would disregard it whenever the country of the habitual residence of the characteristic performer differs from the country where the characteristic performance actually takes place.3
The most contentious scenario to which the escape clause has been applied (indeed, excessively flexibly) by English courts is when the contract at issue is part of a larger chain of transactions between multiple parties and jurisdictions. Bank of Baroda v Vysya Bank Ltd  2 Lloyd’s Rep 87 concerned letters of credit which characteristically involve four parties – the issuing bank, confirming bank, beneficiary and debtor – each having individual bilateral contracts among themselves. As a matter of policy, it is preferable that they are all governed by the same law. Guided by this line of reasoning, Mance J displaced the Article 4(2) presumption as regards the contract between the beneficiary and the issuing bank, finding that the contract was more closely connected with England than India because the letter was issued through National Westminster Bank in London.
Manuel Penadés Fons has defended the English judges’ approach in such chains of contracts scenarios, arguing that they are merely responding to commercial realities and aiming to give commercially sound decisions.4 In support, he refers to Recitals 20 and 21 of Rome I and Article 4(1)(h) which recognise the practicality of there being a single law governing all contracts within the same complex multilateral transaction.5 Thus, English courts’ approach to Article 4 is in his view ‘nuanced’ and context-dependent in a way which does not contravene the Rome Convention, and will not contravene the Rome I Regulation.6 Although the added qualifier ‘manifestly’ in Article 4(3) Rome I suggests an intention to limit the opportunistic exercise of judicial discretion, context will always be everything when there is no definitive guidance on the application of the escape clause.7
However, there remains a conceptual difficulty in determining which contractual relationship to examine first in the search for a single applicable law. The arbitrariness of this choice was evident in Bank of Baroda where the court began its analysis from the contract between the issuing bank and the confirming bank which pointed to England. As Hill suggests in his evaluation of Mance J’s reasoning, it is in fact the contract between the issuing bank and the beneficiary which should be considered first since the beneficiary’s acceptance of it sets in motion the chain of all subsequent transactions.8 The Commercial Court in Molton Street Capital LLP v Shooters Hill Capital Partners LLP  EWHC 3419 – the first English case on Article 4 Rome I – seems to have recognised this when it reasoned that the contracts above and below a contract for the sale of junk bonds were not connecting factors strong enough to displace the Article 4(2) presumption. This most recent English judgement is in line with the new wording of Article 4(3) Rome I in that it reflects the principle of exceptionality enshrined in the word ‘manifestly’. It remains to be seen whether or not it will mark the start of a stricter usage of the escape clause by English judges.
The architectural changes effected by the Rome I Regulation to the European rules for choice of law in contract place an enhanced emphasis on the goal of predictability at the expense of flexibility. English courts – although having often been excessively opportunistic in their use of the choice of law gateways under the Rome Convention – have recently started exhibiting a trend of restraint more in line with the new Rome I provisions, and exercising their discretion more sparingly and conscientiously. Whether this trend will continue is the subject of speculation.
 Jonathan Hill, ‘Choice of Law in Contract Under the Rome Convention: The Approach of the UK Courts’ (2004) 53 ICLQ 325, 329.
 ibid, 342.
 ibid, 343.
 Manuel Penadés Fons, ‘Commercial Choice of Law in Context: Looking Beyond Rome’ (2015) 78(2) MLR 241.
 ibid, 256.
 ibid, 243.
 ibid, 246, 274-276.
 Hill (n 1) 341.