Brussels I Review – Jonathan Hill


Jonathan Hill is Professor of Law at the University of Bristol. He is the author of Cross-Border Consumer Contracts (OUP 2008), The Conflict of Laws (with CMV Clarkson, 3rd edn, OUP 2006), International Commercial Disputes in English Courts (Hart 2005) and is a former editor of Dicey.

Comments on the Review of the Brussels I Regulation

Those who have an interest in private international law (PIL) in Europe have been presented with a valuable opportunity to offer their thoughts on how the Brussels I Regulation should evolve. It has been obvious for many years (indeed, in relation to certain issues, for decades) that the Brussels system is subject to certain weaknesses. At last, there is a chance that (some of) these weaknesses may be addressed.

I have read Andrew Dickinson’s posts with interest and I do not intend to comment on every point which he makes or to offer my own personal answer to every question which the Commission has posed in its Green Paper. Before turning to some of the specific questions on which the Commission is consulting, I have a couple of general observations.

First, Andrew has drawn attention to the unsatisfactory nature of the some of the ECJ’s jurisprudence in the context of the Brussels Convention/ Brussels I Regulation and the need for institutional reform. I suspect that even the ECJ’s greatest supporters would not try to argue that the ECJ has always covered itself in glory when considering the provisions of the Convention/Regulation. My own feeling is that some criticism has been somewhat exaggerated and has not sufficiently acknowledged that the Court’s room for manoeuvre is restricted by a legal text which does not say (and, frequently, cannot plausibly be twisted to say) what one wants it to say. Nevertheless, the PIL community is entitled to better than the fare which has been served up by the ECJ in recent years. The suggestion that, within the ECJ, there should be established a specialist chamber (of PIL experts) to deal with references under the Brussels I Regulation (and other PIL instruments) has been knocking around for well over 30 years. Such reform is seriously overdue.

Secondly, the goal of promoting the ‘good functioning of the internal market’ inevitably provides the backdrop to much of the Commission’s discussion. From the perspective of PIL, this focus runs the risk of distorting priorities. What I would like to see is a principled system of PIL rules which will serve the collective interests of the international litigation community; whether or not this advances the internal market is not my primary concern. So, from my perspective, a rule which arguably has the effect of strengthening the internal market (for example, by simplifying the enforcement of judgments granted against defendants domiciled in a third state) is still a bad rule if it unjustifiably discriminates against non-EU defendants.

The wider international picture

1. One of the most unattractive features of the Regulation is the fact that a judgment granted in one member state against a third state defendant is entitled to recognition and enforcement in other member states, regardless of the basis on which the court of origin assumed jurisdiction. In terms of principle, this approach is indefensible. At the jurisdictional stage, the protection against exorbitant jurisdiction rules which the Regulation offers to EU defendants is not extended to third state defendants; but, at the enforcement stage, non-EU defendants are, nevertheless, exposed to the principle of full faith and credit.

One possible solution is to extend the rules of special jurisdiction in arts 5 and 6 to defendants not domiciled in a member state. Andrew suggests that such extension should not, however, prejudice the application of art 4(1). I am not opposed to Andrew’s suggestion – but I think that any retention of art 4(1) should be subject to a qualification. As regards a defendant not domiciled in a member state, recognition and enforcement under Chapter 3 should depend on the court of origin having assumed jurisdiction on a Regulation basis – or in circumstances in which, had the defendant been domiciled in a member state, the court of origin would have been entitled to assume jurisdiction under the Regulation.

2. Should the Brussels I Regulation be extended to cover the recognition/enforcement of third state judgments? I do not think that there is a compelling case for it to do so. There is no obvious community interest in seeking to determine the circumstances in which a New York judgment is enforceable in England (or France or any other member state). It is imperative that the Community legislator takes seriously the limits of its legislative competence.

3. There is one area involving the relationship between member states and non-member states which needs attention. Whereas art 34(4) deals with the potential problems of conflicting judgments, the Regulation’s silence on potential jurisdictional conflicts between member states and third states is a significant omission. Whatever solution the ECJ might come to in the Goshawk reference, and notwithstanding the arguments surrounding the theory (or theories) of the ‘reflexive effect’ of arts 22, 23, 27 and 28, there is a good case for including within the Brussels I Regulation rules which make provision for proceedings to be stayed or jurisdiction to be declined in cases involving a relevant connection with a non-member state (such as cases where there is a jurisdiction clause in favour of a third state). Some indication of what such rules might look like has been suggested by the European Group for Private International Law (EGPIL). (See arts 22bis, 23bis and 30bis of EGPIL’s Proposed Amendment of Regulation 44/2001 in Order to Apply it to External Situations. While I would not necessarily want to commit myself to EGPIL’s proposed text, EGPIL’s basic approach strikes me as the most plausible solution to the problems posed by the Court of Appeal’s second question in Owusu (ie, the question that the ECJ declined to answer in that case).


In principle, there is a lot to be said for Article 1(2)(d) in its current version. The idea that ‘arbitration’ should be excluded in its entirety from the Brussels I Regulation is intuitively attractive as it marks out arbitration as a field of dispute resolution which is separate from litigation. Of course, there is an interface (court proceedings which relate to arbitration) and the ECJ’s rulings in Van Uden and West Tankers muddy the waters to such an extent that it is essential that the whole question of the relationship between the Regulation and arbitration is revisited. Doing nothing in this area is not a realistic option.

From the jurisdictional point of view, various elements are required. First, the arbitration exception should be removed. Secondly, there needs to be a new rule in Article 22 which, as regards court proceedings relating to arbitration, confers exclusive jurisdiction on the courts of the (putative) seat of arbitration. Thirdly, there is a good case for extending the approach of art 27 to arbitration proceedings. So, if C refers a dispute to arbitration and D initiates court proceedings, the court (which is second seised) should automatically stay its proceedings (without embarking on an investigation of whether the alleged arbitration agreement is valid or not) and, then, if the arbitral tribunal determines that it does have jurisdiction under the arbitration agreement, decline jurisdiction.

In terms of the recognition/enforcement of judgments, a provision dealing with the potential conflict between judgments and awards – along the lines of art 34(4) – would be beneficial. The problem posed by cases where the court of origin wrongly assumes jurisdiction notwithstanding a binding dispute resolution agreement should be addressed. Art 35(1) needs to be amended to allow a defence to recognition/enforcement along the lines of section 32 of the Civil Jurisdiction and Judgments Act 1982. Where the court of origin wrongly assumes jurisdiction in defiance of a valid arbitration clause, the ensuing judgment should not normally be given effect outside the country of origin. In terms of PIL’s priorities, upholding the integrity of dispute resolution agreements (by denying cross-border recognition/enforcement of judgments granted by a non-contractual forum) should be a higher priority than promoting the free flow of judgments regardless of the legitimacy of the assumption of jurisdiction by the court of origin.

Choice of court agreements, lis pendens and related actions

The foregoing paragraph runs in parallel with Andrew’s succinct summary of what is currently wrong under the Brussels I Regulation (as interpreted by the ECJ) with regard to choice of court agreements. The problems surrounding the Gasser decision are well known and there seems to be widespread agreement that its effects need to be reversed. Giving priority to the (putative) contractual forum (and strengthening the effect of jurisdiction agreements by amending the defences to recognition/enforcement) seems the most sensible way forward.

Provisional measures

I agree with the majority of Andrew’s post on this topic. A court seised of substantive proceedings has jurisdiction to grant, in the context of those proceedings, whatever provisional measures are available under its procedural law and art 31 is irrelevant. Where, however, under art 31 court B is acting in support of substantive proceedings brought (or to be brought) in another member state (in court A), one has to accept that court A is the primary court and court B is the secondary court. The ‘real connecting link’ requirement of Van Uden has to be understood in that context. While I agree that the Van Uden requirement is not easy to interpret and apply, there must be limits on what court B can do by way of granting provisional measures of support and some mechanism is required to enable those limits to be set.

In view of the fact that the purpose of art 31 is to allow the granting of measures of support, it makes sense to allow the primary court to decide whether or not the measures granted by the secondary court really are supportive or not. In a situation where the rationale for the grant of a provisional measure is to assist the primary court, how can it be said that it would unduly impinge on national judicial sovereignty to allow the primary court to modify or discharge such a measure if the primary court considers it unhelpful? As things currently stand, a court which, although well-intentioned, is insensitive to (or ignorant of) the system of civil procedure adopted by the primary court may grant provisional measures under art 31 which the primary court considers inappropriate or unduly intrusive. The simplest and most efficient way of counteracting such ‘unhelpful’ support – and promoting better cross-border judicial co-ordination – is to allow the primary court to ‘correct’ the situation by modifying such measures. If this solution were adopted, there would be no need for the ‘real connecting link’ requirement: the secondary court could grant whatever measures it thought would be helpful; the primary court could modify or discharge those measures which it did not consider to be so.