法学（法律）相关参考资料（教案）合集：对于国际贸易仲裁中的暂定保护措施的新方法

New Solutions for Interim Measures of Protection in International Commercial Arbitration: English, German and Hong Kong Law Compared Jan K. Schaefer (National University of Singapore) Readers are reminded that this work is protected by copyright. While they are free to use the ideas expressed in it, they may not copy, distribute or publish the work or part of it, in any form, printed, electronic or otherwise, except for reasonable quoting, clearly indicating the source. Readers are permitted to make copies, electronically or printed, for personal and classroom use. Abstract The new English, German and Hong Kong arbitration legislation has prepared the ground for shifting interim relief to the realm of arbitration. All legal frameworks now address the three main issues that call for solutions in the lex arbitri: the relationship between the courts and arbitration, the arbitrator=s competence to grant interim relief and the enforcement of arbitrator-granted interim relief. Two different concepts can be identified underlying the new laws: on the one hand, there is the court-subsidiarity model of England and, on the other hand, the free-choice model of the Model Law, which was adopted and refined by Germany and Hong Kong. England was the first country to adopt an elaborate court-subsidiarity model; Germany the first country to provide for the enforcement of arbitrator-granted interim relief, even if the seat of arbitration is outside Germany. German law gives a unilateral answer to the question of cross-border enforcement of arbitrator-granted relief - a problem that has not yet received a solution in any international enforcement instrument. The New York Convention does not apply to interim relief. This article introduces the three main issues in a general manner, describes the new solutions of English, German and Hong Kong law in country reports and discusses the different solutions provided against this background on a comparative basis. The court-subsidiarity model will be criticized and the free-choice model recommended. Keywords Arbitration Act 1996; Hong Kong Arbitration Ordinance; German Code of Civil Procedure; UNCITRAL Model Law; lex arbitri; interim measures of protection; interim relief; provisional remedies; access to court; arbitrator=s competence to grant interim relief; enforcement of arbitrator-granted interim relief; cross-border enforcement; party autonomy; court-subsidiarity model; free-choice model; mandatory provisions; arbitration rules. Contents 1. Introduction 2. On the need for interim measures of protection in international commercial arbitration 3. The three main issues 3.1 The relationship between court-ordered interim relief and arbitration 3.1.1 Underlying concepts 3.1.2 Access to the courts 3.2 The arbitrator=s competence to order interim measures of protection 3.3 The enforcement of arbitrator-granted remedies 4. Country reports 4.1 England 4.1.1 Background 4.1.2 Survey of the solutions for the issues 4.1.2.1 The relationship between court-ordered interim relief and arbitration 4.1.2.2 The arbitrator=s competence to order interim measures of protection 4.1.2.3 The enforcement of arbitrator-granted remedies 4.1.3 On the restrictions of party=s choice as prescribed in the mandatory provisions of the lex arbitri 4.1.4 Summary 4.2 Germany 4.2.1 Background 4.2.2 Survey of the solutions for the issues 4.2.2.1 The relationship between court-ordered interim relief and arbitration 4.2.2.2 The arbitrator=s competence to order interim measures of protection 4.2.2.3 The enforcement of arbitrator-granted remedies 4.2.3 On the restrictions of party=s choice as prescribed in the mandatory provisions of the lex arbitri 4.2.4 Summary 4.3 Hong Kong 4.3.1 Background 4.3.2 Survey of the solutions for the issues 4.3.2.1 The relationship between court-ordered interim relief and arbitration 4.3.2.2 The arbitrator=s competence to order interim measures of protection 4.3.2.3 The enforcement of arbitrator-granted remedies 4.3.3 On the restrictions of party=s choice as prescribed in the mandatory provisions of the lex arbitri 4.3.4 Summary 5. Comparative aspects 5.1 The relationship between court-ordered interim relief and arbitration 5.1.1 Underlying concepts compared 5.1.2 Access to the court 5.2 The arbitrator=s competence to order interim measures of protection 5.3 The enforcement of arbitrator-granted interim relief 5.4 On the restrictions of party=s choice as prescribed in the mandatory provisions of the lex arbitri 6. Conclusion 1. Introduction The last few years have seen an increasing activity by legislators to improve the legal framework for arbitration, both domestic and international. More than 30 countries all around the globe have reshaped their arbitration laws since the launch of the UNCITRAL Model Law on International Commercial Arbitration (henceforth Model Law) in 1985. And yet, there are many more to follow. Among the latest legislations are English, German and Hong Kong laws. The Arbitration Act 1996, applicable in England, Wales and Northern Ireland, came into force on 31 January 1997, the Arbitration Ordinance of Hong Kong as amended by the Arbitration (Amendment) Ordinance 1996 on 27 June 1997 and the provisions on arbitration of the German Code of Civil Procedure on 1 January 1998. With regard to interim measures of protection they provide for solutions that go beyond any previous ones in sophistication and completeness. The critical question with regard to interim relief in arbitration is: Who provides interim measures of protection? Shall it be the courts, the arbitrators or both? The answers given in national arbitration legislation and in arbitration rules have changed over the years. Some time ago it seemed to be a common understanding that only courts provide any provisional relief. This was reflected in international instruments such as the >1961 European Convention on International Commercial Arbitration=. Its article VI, paragraph 4 states that a request for interim measures to the courts is not a waiver of the arbitration agreement. Similar provisions are found in arbitration rules. They ensure that a party can have recourse to the courts without fearing to chance the track of dispute settlement by making the court application. No mention was made of an arbitrator=s competence to grant interim measures of protection. But a trend in favour of such an arbitrator=s competence emerged. It was first reflected in arbitration rules such as the >1976 UNCITRAL Arbitration Rules for International Commercial Arbitration= (henceforth UNCITRAL Rules), which provide for a choice of application. Article 26, paragraph 3 of the UNCITRAL Rules refers to court applications and deems them to be compatible with the arbitration agreement. This reiterates the established view. But in paragraphs 1 and 2 of the article, the UNCITRAL Rules go further when making clear that arbitrators have contractual power to order certain kinds of interim measures such as the sale of perishable goods. The arbitrators= order can be established in the form of an interim award under the rules. This prepared new ground and already addressed the three main issues that are at stake when interim measures of protection in arbitration are considered: the relationship between court-ordered interim relief and arbitration, the competence of the arbitrator to order such measures and the enforcement of the arbitrator=s orders. It is necessary to provide solutions for these issues in the lex arbitri as the national legislation plays the decisive part in making provisional remedies in the sphere of arbitration effective. It is the national legislator who provides for court assistance, a suitable legal framework with fall-back provisions and lays down the preconditions for the enforcement of arbitrator-granted interim measures of protection. Neither the New York Convention of 10 June 1958 on the Recognition and Enforcement of Foreign Arbitral Awards (henceforth New York Convention) nor any other international instrument deals with interim measures of protection granted by the arbitrator or their enforcement. The key function of the lex arbitri with regard to interim measures of protection stimulates a comparison of legislations. A comparative synopsis of different national solutions can highlight their strengths and weaknesses. It can also further the discussion about the most suitable solution for international commercial arbitration. As legislators are competing to provide the most favourable framework for international commercial arbitration B they either want to defend their position as an international arbitration venue (e.g. England and Hong Kong), establish their country as such a venue (e.g. Germany) or signal a positive investment climate (e.g. India) - there is a real chance that inadequacies can and will be remedied in future legislation. The arbitration laws of the three countries compared follow different approaches. Each approach is worked out in a very elaborate and sophisticated manner. They represent alternative role models for future legislators. English law, on the one hand, provides an approach that can be called a court-subsidiarity model. Provisional remedies should in the first place be applied for before the arbitrator. Court intervention is the last resort. The court=s jurisdiction is restricted; it depends on the arbitrator=s power to act effectively and is therefore subsidiary to it. This approach shifts interim measures as far as possible to the realm of arbitration. The English legislator is the first one to adopt this model. Parties can only opt out of it if they do not empower the arbitrator to grant interim measures of protection. German law, on the other hand, follows the free-choice approach of the Model Law. Arbitration parties are free to apply either to the court or to an arbitrator. Legislation provides for mechanisms that make arbitrator-granted relief as far as possible equally effective as court-ordered relief. It provides a real alternative for the parties. Hong Kong law adopted an approach that is in between the free-choice and the court-subsidiarity approaches. The key lies with the courts to determine with which model Hong Kong will align itself. Germany is the first country which provided for cross-border enforcement of arbitrator-granted interim relief in its national law. Under German law, arbitrator-granted interim relief can be enforced in Germany even if the seat of arbitration is outside Germany. This is a progressive, unilateral step to solve the cross-border enforcement issue especially relevant with respect to international commercial arbitrations. If all countries follow this example, the need for an international instrument to provide a cross-border enforcement mechanism for interim measures of protection can be relieved. This article presents the new laws, examines their differences and evaluates them against a background of a general discussion of the three main issues involved. 2. On the need for interim measures of protection in international commercial arbitration Today, interim measures of protection form part of the regular litigation process. Several reasons contribute to this. The main reason can be found in the long duration of court proceedings that call for interim solutions. But it is not only with litigation that the time span between the beginning of the dispute and its resolution has increased in recent years; the same applies for international commercial arbitration. According to Craig, Park and Paulsson (1990, pp. 20-21), the average duration of International Chamber of Commerce (ICC) arbitrations is between one and two years. The main reasons for the long duration of international commercial arbitrations lie in the special circumstances of the process. Among the special circumstances are the frequently great geographical distances between the dramatis personae and the coordination of the busy schedules of international arbitrators and party counsels. But the delay cannot only be attributed to inherent causes; it also finds its ground in the dilatory tactics applied by a party to which arbitration is not immune (Knoepfler, 1997, p. 307). Thus, the need for interim solutions arises in arbitration as it does in commercial litigation. Justice can only be done if efficient interim relief is available. Provisional remedies come under the principle that >justice is not to be evaded= (Andrews, 1994, p. 20). The means of recourse differ from country to country, but the system itself is regarded as >one of those general principles of law common to all legal systems= (Collins, 1994, p. 10). Interim relief, which has not received a legal definition, can be grouped as follows: To preserve the status quo in order to ensure enforcement, to shape the relationship between the parties during the process of dispute settlement and to preserve evidence (see Knoepfler and Schweizer, 1984, pp. 223-224, with reference to German doctrine). It is hardly possible to make general remarks as to the kind of provisional remedies that is likely to be needed for what kind of legal relationships. It depends on the special circumstances of each individual case. Thus it is important to have the whole armoury of interim measures at hand once the need for them arises. The impact of interim relief for the well-functioning of any method of adjudication can be illustrated with regard to Mareva injunctions (after The Mareva). Mareva injunctions prevent the dissipation of assets. If such a mechanism would not be available for the time that passes until final adjudication takes place, the winning party would only obtain a >Pyrrhic victory= (Van den Berg, 1981, p. 143). 3. The three main issues The three main issues are the relationship between court-ordered interim relief and arbitration, the competence of the arbitrator to order such measures and the enforcement of the arbitrator=s orders. The following exposition focuses on theoretical aspects of the three issues and discusses them broadly. 3.1 The relationship between court-ordered interim relief and arbitration 3.1.1 Underlying concepts Interim relief in arbitration is an interface between private dispute settlement and the court. It is one of these aspects of arbitration procedure that cannot escape court interference. The arbitrator has no power to enforce his orders. As the effectiveness of an interim measure of protection depends, in the end, on its enforceability, court support may be needed. The critical question is how best to shift interim measures of protection with their need for enforcement sanctions to the realm of arbitration. Theoretically, three basic possibilities can be identified. The first one is that granting interim measures is exclusively allocated to the court. The court would provide the same interim protection to arbitration parties as it does to litigation parties. The opposite extreme is that one shifts interim measures of protection exclusively to the sphere of arbitration and only leaves the enforcement of the arbitrator=s orders to the courts. This would mirror the regular arbitration process. The arbitrator decides the dispute in his award and the courts enforce that award. The courts do not interfere with the decision of the arbitrator. Finally, there is the option of free access to both the court and the arbitrator for interim relief. The latter two possibilities call for a legal infrastructure. It must be ensured that the arbitrator can effectively grant interim measures of protection. This requires that his competence is spelled out and that an enforcement mechanism is adopted that suits the special needs of interim measures of protection. The measures must be enforceable very speedily and, where necessary, on an ex parte basis to preserve the element of surprise. Today, a small number of countries follow the first possibility when they rule out any arbitrator=s competence to grant interim measures of protection. Examples are Italy and Greece (Sanders, 1996, p. 113). The majority follow the last approach as suggested by the Model Law when they provide for court access and the arbitrator=s competence to order interim measures of protection. But only spelling out the arbitrator=s power is not enough to provide a real alternative: a suitable enforcement mechanism must accompany it. The Model Law did not provide a guideline for an enforcement mechanism. Most countries do not provide a truly suitable mechanism. Germany and Hong Kong have implemented such a mechanism. No country has yet adopted the second possibility. The English court-subsidiarity model comes close to it in defining preconditions for court access. The arbitrator is the course of first resort; the court is the course of last resort with regard to interim measures. It is a concept favoured and coined by German authors as the court-subsidiarity model (see Schlosser, 1989, pp. 306-307). Nevertheless, the new German law follows the free-choice approach of the Model Law. But a subsidiarity model is reflected in the approach the ICC fosters. Article 23 (2) of the ICC Rules (in force as from 1 January 1998) states that an arbitration party can only in >appropriate circumstances= apply to the courts for interim measures of protection once the file has been transmitted to the arbitral tribunal. With the >appropriate= test, the ICC Rules set up a hurdle that needs to be taken before a party can obtain court protection. This mechanism mirrors what can be regarded as a policy of court subsidiarity. However, it should be noted that article 23 (2) ICC Rules is wider than its predecessor, article 8 (5). Article 8 (5) referred to >exceptional circumstances=. 3.1.2 Access to the courts If the lex arbitri does not clearly state that a party can have recourse to the courts for orders of interim protection, a dispute may arise as to whether seeking such recourse to the courts constitutes a waiver of the arbitration agreement. But it can also have as a consequence that the courts refuse to aid arbitration parties, as has been the case in New York. In the McCreary decision, the Third Circuit held that the New York Convention >forbids the courts of a contracting state from entertaining a suit which violates an agreement to arbitrate. Thus the contention that arbitration is merely another method of trial, to which state provisional remedies should equally apply, is unavailable [.?.?.]=. The reasoning of McCreary is, however, not followed by all American courts. In Carolina Power & Light Company v. Uranex it was held: >There is no indication in either the text or the apparent policies of the [New York] Convention that resort to prejudgment attachment was to be precluded [.?.?.].= As the Supreme Court had not yet an opportunity to rule on the question, American jurisprudence differs on the availability of interim measures of protection from the courts (see Born, 1994, pp. 772-773). A provision providing for both access to courts and the compatibility of court-ordered interim relief with arbitration is essential to do justice to the arbitration parties= cause and to prevent uncertainty. If an arbitrator could equally efficiently order all measures of interim relief that the court provides, there would hardly arise any need and be any justification for court applications by arbitration parties. But as the arbitrator=s jurisdiction is limited, the court=s assistance is needed. The first limitation of the arbitrator=s jurisdiction is inherent in the arbitral process. The arbitrator derives his power from the arbitration agreement. The arbitration agreement is a contract that only binds the parties who entered into it. It has no effect on third parties. A Mareva injunction ordered by an arbitrator could thus only bind the arbitration parties, not their banks where the assets are held. Secondly, an arbitrator lacks any enforcement power. As a consequence, the Mareva injunction ordered by him would not be fortified with the threat of contempt of court, but would only have contractual effect between the parties. Taking these two limitations into account, it becomes evident how important access to the courts is for certain kinds of interim relief in arbitration. Even if enforcement of arbitrator-granted interim relief is provided for, there will always be one step added to the procedure if enforcement is at stake. A court needs to step in to grant permission to enforce the arbitrator-granted order. This causes delay. But it is not only the need for efficiency and speed that demands access to the courts for arbitration parties, it is also administrative peculiarities of arbitration. There is no permanent arbitral tribunal to which international commercial arbitrations are submitted. A tribunal has to be established for each requested arbitration. This takes time. As long as the file has not been transferred to the arbitrator, there is no one available to grant interim measures of protection. Litigation practice shows that interim measures of protection are regularly in the highest demand before the case proceeds to trial. This observation is equally valid for arbitration. The greatest need for provisional remedies arises at a point when the tribunal has not yet been established. Arbitral institutions try to remedy this problem. An attempt undertaken by the International Chamber of Commerce in 1990 with its >Pre-Arbitral Referee Procedure= is, however, not successful (Schwartz, 1993, p. 64). Other institutions such as the World Intellectual Property Organization (WIPO) with its progressive draft >WIPO Emergency Relief Rules= look for alternatives. The London Court of International Arbitration (LCIA) dropped its proposal in 1997. The above shows how essential court involvement is under certain circumstances. But there are also voices against it. For instance, Boesch (1989, p. 8) warns vehemently about the risks an application for interim relief to the courts can carry for arbitration. There are indeed risks involved, especially if the application to the court is regarded as a dilatory tactic, a way to get access to a biased forum or as a means to put pressure on the other party. Threatening to go to a public court with a dispute that should be solved confidentially can do this. But such threats come from a party that does not act in good faith. This kind of behaviour should not be taken as a reason to bar a good-faith party from access to the most efficient remedy, especially if efficiency is decisive to secure assets without which the whole process of dispute settlement is rendered superfluous. Arbitration is there to serve the rights of the parties and is not an end in itself that needs to be protected at the expense of the bona fide party against attacks from the mala fide party. There are other means to prevent such attacks, for instance by stipulating that provisional remedies from courts in certain countries - the courts of which are deemed to be biased or bribable B are unavailable. Whether the courts will recognize such an opting-out of court protection is, however, doubtful: English courts do (section 44, subsection 1 of the Arbitration Act 1996), German ones do not (LG Frankfurt). Further, it could be stipulated contractually that a party needs the arbitrator=s permission to apply to the courts. This can be regarded as a means to filter mala fide applications. But it could give rise to dispute concerning the stipulation and thereby hamper the whole process. The task to dismiss prima facie dilatory or tactical applications is best entrusted to experienced judges without imposing restraints on court access by the parties. The relationship between courts and arbitration should be one of mutual trust, respect and support. Modern trend is gradually moving in this direction. A centralization of arbitration matters at one court - preferably a hierarchically higher one - would serve this goal. 3.2 The arbitrator=s competence to order interim measures of protection The lex arbitri should also provide for the arbitrator=s competence to order interim measures of protection. It is not only necessary to indicate that the state recognizes such a power, which is not yet the case in all countries, such as in Italy (article 818 Code of Civil Procedure), but it is also helpful to give the arbitrator explicit assurance of his power. This will especially be the case in institutional arbitrations because the contractual power conferred to them in the arbitration rules (e.g. article 23, paragraph 1 of the 1998 ICC Rules or rule 25 (g) and rule 25 (j) of the 1997 Arbitration Rules of the Singapore International Arbitration Centre - SIAC) is backed up in the lex arbitri of the seat of arbitration. Then there is no risk involved that the award will be set aside or will not be enforced on the ground that the >arbitral procedure [.?.?.] was not in accordance with the law of the country where the arbitration took place= (article V, paragraph 1 (d) of the New York Convention for non-enforcement ground; compare article 34, paragraph 2 (a) (iv) of the UNCITRAL Model Law for setting aside ground). Arbitrators who are reluctant to grant provisional remedies might find relief in such an assurance. 3.3 The enforcement of arbitrator-granted remedies An arbitrator has no imperium, >the right [.?.?.] to enforce the laws= (Black, 1891, p. 594). Thus the state needs to step in with its enforcement machinery to lend the arbitrators= order the necessary authority to ensure compliance. Whether the enforcement machinery will be invoked is a different matter. Indeed Craig, Park and Paulsson point out that >parties do not ordinarily flout procedural orders made by arbitrators under contractually granted powers= (1990, pp. 145-146). But as the inclusion of >ordinarily= implies, there are situations in which enforcement is necessary to ensure compliance. In order to be a real alternative to court-ordered interim relief, the orders of the arbitrators must be fortified with a real threat of enforcement. To achieve this end, state support is needed. Cross-border enforcement of arbitrator-granted interim relief is an important aspect of the enforcement issue with regard to international commercial arbitration. In the absence of an international regime for cross-border enforcement, a unilateral approach can be seen as a way to solve the problem. The New York Convention is silent on the question of interim measures of protection and their cross-border enforcement (Van den Berg, 1981, p. 144). The Supreme Court of Queensland (see Sanders, 1996, p. 115) rejected an attempt to enforce an interim award under the New York Convention in 1993. Berger (1993, p. 345 with further reference) outlines the convincing arguments against the enforcement of interim measures of protection under the New York Convention. In a cross-border context, arbitrator-granted interim relief could become even more efficient than court-ordered relief if at least a unilateral enforcement regime is in place. The importance of cross-border enforcement will be illustrated below with regard to interim measures preventing the dissipation of assets. An arbitration will often take place in a third, >neutral= country. This is a place where substantive assets of the parties are most likely not held. If an arbitrator orders interim relief, it is (with the exception of German law) - at best - only enforceable at the seat of arbitration. The effect of the provisional remedies is therefore rather limited. It will not reach assets in the home countries of the parties. If a court at the seat of arbitration grants an interim order, it can only prevent the dissipation of assets out of its jurisdiction. In the absence of a bilateral or multilateral enforcement agreement covering court-ordered or arbitrator-granted interim relief, the jurisdictional limitations force an arbitration party to apply to courts in the country or even countries where the assets are held. This is the reason why it is so important to provide for court access even if the seat of arbitration is outside the country where the order is sought. Article 1(2) of the Model Law caters for this need. If interim relief in arbitration would be cross-border enforceable, one order would be sufficient and no need would arise to apply to several courts for an order under probably different regimes of interim relief. Only enforcement proceedings in different countries would be needed. It is unlikely that a country enforces interim relief granted by a court of a foreign country without an agreement providing for reciprocity. To expect a global, multilateral enforcement agreement for court-ordered interim relief is illusionary. To establish an international regime for the cross-border enforcement of arbitrator-granted interim relief will be difficult to achieve. A possibility would be to amend the New York Convention. In the meantime, unilateral solutions could provide for cross-border effectiveness in the sense that countries render their mechanisms for enforcing arbitrator-granted interim relief not only applicable if they are chosen as arbitration venue but also in the case that the seat of arbitration is somewhere outside the country. This would reflect the legislator=s commitment to the cause of arbitration as a whole. 4. Country reports 4.1 England 4.1.1 Background England is an established international arbitration venue, for both institutional and ad hoc arbitrations. The London Court of International Arbitration is one of the busiest arbitration institutions in the world. It plays in the league of the market leader, the Paris-based International Chamber of Commerce, the American Arbitration Association (AAA) and the traditional arbitration centres in Vienna, Stockholm and Zurich. Prior to the new single Arbitration Act 1996, the English legal framework for international commercial arbitration comprised not only three different Acts (Arbitration Acts 1950, 1975 and 1979) but also a voluminous body of case law. The diversity of sources made access to the law equally difficult for non-specialists as foreigners. It therefore ill suited international commercial arbitration with its need for a transparent legal regime. But it was not only the law=s presentation which did not keep pace with modern trends, also substantive matters called for reform to defend England=s market share in the arbitration business. The Model Law set new standards when it was adopted by UNCITRAL and recommended by the General Assembly of the United Nations in 1985. It gave impetus to a >Departmental Advisory Committee= (DAC), chaired by Mustill L.J. (as he then was). The task of the Committee was to advise on the adoption of the Model Law in England. The committee rejected adoption on several grounds, among them being the Committee=s finding that the Model Law differed >from established principles of English law= (see Merkin, 1996, p. 2, for quotation from the 1989 DAC Report). Even though the committee rejected adoption of the Model Law, it urged for new legislation and recommended features thereof. One of the features was the restatement of the principles of English arbitration law. But >consideration should also be given [that the new statute] should, so far as possible, have the same structure and language as the Model Law, so as to enhance its accessibility to those who are familiar with the Model Law= (see DAC Report, 1996, No. 2 for quotation of paragraph 108 (7) of the 1989 DAC Report). The first bill published in 1994 continued to reject adoption of the Model Law, as did the second bill prepared by the DAC in 1995 - chaired by Saville L.J. - which followed the abandonment of the 1994 draft. The 1995 draft led, with modifications, to the enactment of the Arbitration Act 1996. The >DAC Report on the Arbitration Bill= of February 1996 contains a comprehensive commentary on the provisions and points out that >very close regard was paid to the Model Law, and [that] it will be seen that both the structure and the content of [.?.?.] the final draft owe much to this model= (DAC Report, 1996, No. 4). The Arbitration Act 1996 came into force on 31 January 1997 (The Arbitration Act 1996 (Commencement No. 1) Order 1996). The new Arbitration Act 1996 was praised in the House of Lords and in the House of Commons (see Miller, 1996, p. 252 and p. 256) and has been welcomed >as it introduces a new arbitration regime which is clear and workable= (Slaughter and May, 1997, p. 1). The regime applies equally for domestic and international arbitrations (see Collins, 1997, pp. 96-97). 4.1.2 Survey of the solutions for the issues 4.1.2.1 The relationship between court-ordered interim relief and arbitration Underlying concept The underlying concept will be referred to as a policy of subsidiary court jurisdiction. The task to grant interim relief is in the first place allocated to the arbitral tribunal. The court can only step in if the arbitrator has no power or is unable for the time being to act effectively. A court-subsidiarity approach requires a conflict rule. It is for the conflict rule to determine the preconditions for court jurisdiction. It is in place to set the limits of court jurisdiction and can therefore also be referred to as limitation rule. In English law it is the >effectiveness test= of section 44 (5) Arbitration Act 1996 (hereafter all provisions without further reference are those of the Arbitration Act 1996). Access to the court The central provision in the Arbitration Act 1996 is section 44. It is about >court powers exercisable in support of arbitral proceedings=. Further relevant provisions are section 105 in connection with >The High Court and County Courts (Allocation of Arbitration Proceedings) Order 1996=, which determines that the High Court has general jurisdiction to grant interim measures of protection under section 44, although the Central London County Court may have limited jurisdiction in small cases under certain circumstances. Section 2 subsection 3 (b) extends the courts jurisdiction. If the seat of the arbitration is overseas or has not been designated or determined, the court has discretion to grant interim relief under section 44. The court=s decision will be guided by whether or not it would be appropriate to do so. Therefore the provision restates the principle as developed in the Channel Tunnel case. The Arbitration Act 1996 further provides for legal definitions of terms used in section 44. A comprehensive list of minor definitions is set out in section 82. Section 83 indexes all defined expressions in a user-friendly manner. Order 73 of the Rules of the Supreme Court (RSC) has been newly drafted and renamed >Applications Relating to Arbitration= by the >Rules of the Supreme Court (Amendment) 1996= (see Practice Note by Colman J., 1997, pp. 379-380). RSC Ord. 73, rule 18 (1) has special relevance for the interpretation of section 44 (3). RSC Ord. 73, rule 15 (2) states that generally all arbitration applications shall be heard in chambers. Section 44 has seven subsections. They lay out a sophisticated mechanism for the relationship between the court and the arbitral tribunal with regard to interim measures of protection. The large number of subsections is due to the detailed elaboration of the subsidiarity approach. Section 44 does not specifically include a provision stating that seeking recourse to the court for interim relief does not constitute a waiver of the arbitration agreement. But the very existence of the section can be regarded as such an intention. In the following lines, each subsection of section 44 will be briefly analysed. Subsection 1 grants the court the same power of making orders about matters specified in subsection 2 as it has for the purpose and in relation to legal proceedings. It also provides for the parties= choice to opt out of court support. The possibility to opt out of court-sponsored interim relief is to be welcomed especially since the parties can rule out court support in certain matters and retain it in others. This right to choose reflects party autonomy. But it should be noted that opting-out of court protection under section 44 will strip the parties of access to ex parte Mareva injunctions and Anton Piller orders (after Anton Piller KG v. Manufacturing Processes Ltd.) as these are remedies an arbitrator cannot grant even if the parties conferred power to him under section 39 (infra). If they further omit to opt in section 39, the scope of interim measures of protection will be reduced to the remedies available under section 38. Among them are security for costs (section 38 (3)). Section 44 (2) lists matters in respect of which jurisdiction has been conferred upon the courts, such as the taking of evidence of witnesses, the preservation of evidence, the sale of any goods that are the subject of the proceedings or the granting of an interim injunction. The issues are addressed in a concrete manner but some of them overlap, which makes it B at first glance B confusing to decide on which matter to base jurisdiction for an interim measure of protection known from litigation. By way of an example, a look at the commentaries on the new Act shows an astonishing diversity of answers to the question of where to ground the court=s jurisdiction for an Anton Piller order. Rutherford and Sims (1996, p. 152) regard section 44 (2) (c) as the basis, whereas Harris, Planterose and Tecks (1996, p. 179) prefer section 44 (2) (b). Merkin (1996, p. 72) seems to locate the jurisdiction to grant Anton Piller orders even in section 44 (3). This is, however, wrong as the court only has power with respect to the matters listed in subsection 2. Thus any relief must have a jurisdictional basis in a matter listed in that subsection. But also the localization by the other authors is only sensible at first glance. In each case, the wording of the two matters chosen matches the aim of an Anton Piller order. Lord Denning MR described it as follows in Yousif v. Salama: >[An Anton Piller order] would be preserving the evidence in the case.= But a systematic look taken at the three norms dealing with jurisdiction for interim measures of protection (sections 38, 39 and 44) reveals that neither of their choices is correct. It becomes clear from section 44 (5) that sections 38, 39 and 44 form a systematic union. Section 44 (5) incorporates the conflict or limitation rule central to the subsidiarity approach. Only if an arbitrator has no power to grant the respective remedy, the court has jurisdiction without passing the >effectiveness= test. For any matter listed in section 38, the >effectiveness= test of section 44 (5) must be passed because the arbitrator possesses the powers described in section 38 ex lege, i.e., they are conferred to him by statute. An exception would be that the parties ruled out a certain power listed in section 38. The consequence is that, if a matter listed in section 44 (2) forms the jurisdictional basis of a specific interim measure of protection granted by the court, the complementary power in section 38 will provide for the arbitrator=s jurisdiction with respect to this remedy. This is the case as otherwise the conflict or limitation rule of section 44 (5) could not fulfill its function. It depends on a system of complementary powers to transparently determine the cases (other than the obvious ones affecting third parties) in which an arbitrator has no power. The authors mentioned above base the court=s jurisdiction for an Anton Piller Order on paragraphs 44 (2) (c ) and (b) respectively, which complement subsections 38 (4) and (6) respectively. This choice implies, as contrasted to the above, that an arbitrator has the power to order an Anton Piller order - not affecting third parties - under section 38 subsection 4 or subsection 6. This, however, contradicts the DAC Report on section 39. The commentary on that section rules out an arbitrator=s power to grant an Anton Piller order (1996, No. 201). But if the parties cannot confer such a power to the arbitrator, how can it have been conferred ex lege in the first place? Therefore it is not enough to look at the wording of the matters mentioned in section 44 (2) alone to decide which matter forms the basis of the court=s jurisdiction for a certain measure of interim protection. It is also necessary to pay attention to the systematic consequences that flow from such an allocation for an arbitrator=s power under section 38. Thus it seems only convincing to regard section 44 (2) (e) as the jurisdictional basis for an Anton Piller order granted by the court. This analysis has shown that the provisions on interim relief have a systematically interlinked structure. The overlap of matters in section 44 (2), especially paragraphs (b), (c) and (e) with respect to an Anton Piller order, does not cloud the clarity of the section as long as a systematic allocation having regard to sections 38 and 39 takes place. This is not only an academic observation, but also has practical implications. RSC Ord. 73, rule 4 (d) requires (where appropriate) the specification of the section under which the application is brought. It may be sufficient to quote section 44 to satisfy this rule, but RSC Ord. 73, rule 18 (1) requires for applications under section 44 that the application states the reasons: >Why the deponent believes that the conditions in section 44 (5) is satisfied.= The most uncontroversial way to do so is by relying on the ground that the arbitrator has no power. This is the case for the Anton Piller order. Only if, according to subsections 1 and 2, the jurisdiction of the court for a certain measure of protection is established, the next subsections become relevant. Subsections 3, 4 and 5 determine whether the court can exercise the established jurisdiction. Subsections 3 and 4 differ between urgent and non-urgent applications. In the case of urgency, a party can apply to the court without permission of the tribunal or without a special, written party agreement. In the case of a non-urgent application, it cannot. The concept of urgency is central to determining the correct application procedure. In subsection 3 special reference is made to orders necessary for the purpose of preserving evidence or assets. This refers to Anton Piller orders and Mareva injunctions. These orders are generally granted ex parte. A comparison of subsections 3 and 4 gives rise to the interpretation that ex parte orders can also be granted by court in urgent situations in support of arbitration. It is not only the special - though indirect - reference to Mareva injunctions and Anton Piller orders which support this interpretation, but especially the wording of section 4. A part put in brackets reads: >Upon notice to the other parties and to the tribunal.= This has a double effect. Firstly, it states that ex parte orders are unavailable in non-urgent cases and, secondly, that they are available in urgent cases (argumentum e contrario). RSC Ord. 73, rule 18 (1) refers to ex parte applications. It thereby supports the above interpretation. Rule 18 (1) lists the reasons an affidavit for an urgent ex parte application shall contain. The >urgency= requirement of section 44 (3) is fulfilled if reason is given >why it was not practical to obtain [the] permission or agreement= (RSC Ord. 73, rule 18 (1) (b)). Urgency is not automatically assumed once an ex parte order is at stake; it depends on the practicability. The concept of >effectiveness= is the key to court assistance. It is central to the subsidiarity approach and is laid down in subsection 5. This subsection reads: >In any case [urgent or non-urgent] the court shall act only if or to the extent that the arbitral tribunal, and any arbitral or other institution or person vested by the parties with power in that regard, has no power or is unable for the time being to act effectively.= The subsection has to be read in connection with sections 38 and 39, as has been argued above. Subsection 5 is especially relevant for institutional arbitrations and ad hoc arbitrations which provide for an arbitrator=s power under section 39 (e.g. if the UNCITRAL Rules are agreed upon (article 26 (1)). In that case, the >effectiveness= test will have to be applied to all court applications. An application for ex parte Mareva injunctions or Anton Piller orders is the exception as for them the arbitrator lacks power (infra). RSC Ord. 73, rule 18 (1) and (2) state that in cases of urgency and non-urgency the application shall state the reasons >why the deponent believes that the condition in section 44 (5) is satisfied=. With ex parte Mareva injunctions and Anton Piller orders the task is simple. But for all other cases the key to court access is the inability of the arbitrator for the time being to act effectively. Thomas (1997, p. 408) writes that >at present [there is] no indication of how the test in section 44 (5) will be applied=. Taking the above theoretical considerations into account, it can be assumed that an arbitrator=s relief is always a step behind the court=s relief in terms of effectiveness because the enforcement of an arbitrator=s order calls for court involvement which adds an additional step to the procedure. Whether that delay amounts to a lack of effectiveness is to be determined by the courts. It is, however, unlikely that they so rule, as this would render the conflict rule obsolete. The subsidiarity model would automatically turn into a free-choice one. Subsection 6 deals with another aspect of the subsidiarity approach. The court can order that its order ceases to have effect once the arbitrator deals with the matter. Finally, subsection 7 states that leave from the court is required for any appeal from a decision of the court under this section. This provision provides a disincentive to parties to dispute over the mechanism of section 44 in order to sabotage the arbitration. Section 44 incorporates the subsidiarity approach in a very stringent manner. It shows clearly that such an approach requires a very sophisticated legislative framework. Such a framework is provided in section 44. The English approach can be regarded as a very thoughtful and systematic example of an elaborate subsidiarity model. 4.1.2.2 The arbitrator=s competence to order interim measures of protection Sections 38 and 39 are the relevant provisions. Some of the powers conferred to the arbitrator under section 38 deal with interim measures of protection such as security for the costs of the arbitration (section 38 (3)) or the preservation of evidence for the purpose of the proceedings (section 38 (6)). But generally section 38 deals with general powers. More specific is section 39, which concerns provisional orders. Even though the latter=s short title refers to >awards=, the term >orders= is to be preferred as the section only refers to orders (Sutton, Kendall and Gill, 1997, p. 257, footnote 48). For the interpretation of section 39 (1) recourse needs to be taken to section 48 (2) and (5). Section 48 deals with the remedies an arbitrator can grant in an award and which section 39 (1) makes available upon party agreement on a provisional basis. The arbitrator=s power to make provisional orders is a novelty in English law. The Kostas Melas was the authority to prohibit an arbitrator to make provisional financial arrangements between the parties. Subsection 1 ensures that the arbitrator=s power rests on party agreement and determines the scope of remedies. It reads: >The parties are free to agree that the tribunal shall have power to order on a provisional basis any relief which it would have power to grant in a final award.= Subsection 2 gives two examples: they are a provisional order for the payment of money or the disposition of property as between the parties and an interim payment on account of the costs of the arbitration. The examples do not mention injunctive relief but are not exhaustive. Thus the interpretation of subsection 1 in connection with section 48 (2) and (5) (a) is decisive in determining whether an arbitrator can grant injunctive relief. Section 48 (5) (a) reads: >The tribunal has the same powers as the court to order a party to do or refrain from doing anything.= The former part refers to mandatory and the latter one to prohibitory injunctions. Both are available as interim injunctions under section 39 (Thomas, 1997, p. 405 and Sutton, Kendall and Gill, 1997, p. 386, footnote 76). Notable exceptions are ex parte Mareva injunctions and Anton Piller orders. However, simply reading section 39 (1) in connection with section 48 (2) and (5) (a) cannot identify these two exceptions. The wording of the sections rather gives rise to the assumption that the two remedies are available from an arbitrator. To find out about the exceptions recourse needs to be taken to the DAC Report. The relevant part of the DAC Report reads: >We should note in passing that the July 1995 draft would arguably (and inadvertently) have allowed arbitrators to order ex parte Mareva injunctions or even Anton Piller relief. These Draconian powers are best left to be applied by the Courts, and the provisions of the Bill with respect to such powers have been adjusted accordingly.= (1996, No. 201) Commentaries on the Act such as Russell on Arbitration (Sutton, Kendall and Gill, 1997, p. 258) plainly state the exceptions with reference to the DAC Report. It remains to be seen how the courts construe the relevant sections. From a practical point of view it is, however, very unlikely that a party applies to an arbitrator to grant an ex parte Mareva injunction or an Anton Piller order. Therefore, the chance of a judicial clarification is rather small. With these restrictions, the scope of arbitrator-granted relief is limited under English law. It is interesting to note that the DAC Report (1996, No. 201) states: >that [.?.?.] a number of practising arbitrators [.?.?.] would be unhappy with [the power to make temporary or provisional financial arrangements between the parties] and saw no need for [it].= But it also mentions (No. 203): >[that the provision] could serve a very useful purpose, for example in trades and industries where cash flow is of particular importance.= Subsection 3 restates the provisional nature of the order where it states that any such order is to be subject to the tribunal=s final adjudication. It further states that the tribunal=s final award is to take account of any provisional order. This ensures that parties do not pay more than is awarded from an overall perspective. Subsection 4 states that the parties need to confer by agreement power to the arbitrator to grant provisional orders. In default of such an agreement the arbitral tribunal has no such power. Section 39 does not affect the power of an arbitrator under section 47 dealing with partial awards. This note serves to clarify the difference between a provisional award (order) and a partial award. Both used to be referred to as interim awards; this blurred the difference. 4.1.2.3 The enforcement of arbitrator-granted remedies English law provides a solution for this issue in section 42. The mechanism adopted under English law reflects the underlying policy of court subsidiarity. The provision complements the subsidiarity concept by giving >teeth= to arbitrator-granted interim measures of protection. It has to be seen in connection with sections 38 and 39. It provides the enforcement mechanism for orders granted under these sections. Neither section 38 nor section 39 refers to an enforcement of the arbitrator=s order in the form of an award under section 66. Thomas (1997, p. 407) expresses the hope that an arbitrator=s order granted under section 39 could also be enforced as an award under section 66. This is, however, doubtful as he concedes himself. Section 42 is likely to prevail over section 66 as the more special rule. Subsection 1 states that the court may make an order requiring a party to comply with a peremptory order made by the tribunal. The court=s order is fortified by contempt of court threat. It can further be assumed that a third party aiding the arbitration party to break the order will be liable for contempt of court (in line with Acrow (Automation) Ltd. v. Rex Chainbelt Inc.). But the arbitrator=s limited jurisdiction with regard to binding third parties remains. It is possible for the parties to opt out of the enforcement mechanism. Subsection 2 deals with the question of who can apply for court support. There are basically three different ways. The application can either be made by the tribunal upon notice to the parties, by an arbitration party with the permission of the tribunal and upon notice to the other party or by a party without permission of the tribunal and without notice to the other party if the parties have agreed that the powers of the court under this section shall be available. When drafting an arbitration agreement this point should be considered. The provisions that reflect the subsidiarity approach are subsections 3 and 4. Subsection 3 states that first of all any arbitral process must be exhausted that could make a party comply with the arbitrator=s order. This refers to section 41, which spells out a whole catalogue of remedies the arbitral tribunal has in the case of a party=s default. The court must be satisfied that the applicant has exhausted every arbitral mechanism available. This prevents a party from applying for court enforcement without an attempt to address non-compliance within the arbitration context. Subsection 4 supports this notion with regard to time. A party cannot seek help from the court until a reasonable time has expired during which the other party is given a chance to comply with the order. These two hurdles on the way to court access ensure that court support is only the last resort. Subsection 5, finally, states that leave of the court for any appeal from a decision of the court under this section is necessary. The mechanism provided for in section 42 cannot enforce urgent measures of protection. It therefore does not suit injunctive relief well. It is more like a reinsurance for the arbitrator. He need not fear the >toothlessness= of his orders. He could easily perceive orders that do not carry any threat of coercive enforcement as a threat to his authority. If section 42 helps to give him confidence, the section would fulfill an important role. Practice will show how the courts interpret the hurdles that need to be taken by an application. Section 42 has to be seen in the context of the subsidiarity approach. If the courts set the standards very high for being satisfied under section 42 (3) and (4), arbitrator-granted interim relief will receive a setback. Parties would then prefer interim relief from the courts and give a second thought about opting in section 39. Whether the courts support the subsidiarity approach underlying the Act or rather obstruct it will be seen in future. The interpretation of section 42 can be seen as an indicator that shows whether the subsidiarity approach lives. English law does not provide for the applicability of section 42 if the seat of the arbitration is outside England, Wales or Northern Ireland. It is therefore not possible to apply for the enforcement of a >foreign= arbitrator-granted interim order under this provision. 4.1.3 On the restrictions of party=s choice as prescribed in the mandatory provisions of the lex arbitri The English Act follows the territoriality principle of the Model Law when it states, in section 2 (1), that the provisions of this Part [part 1] apply where the seat of the arbitration is in England and Wales or Northern Ireland. This means that all mandatory provisions of the Act have to be observed (compare Union of India v. McDonnell Douglas Inc). A list of mandatory provisions is annexed to the Act in Schedule 1. None of sections 38, 39, 41 and 42 nor section 44 qualify as mandatory provisions. The parties are free to depart from non-mandatory provisions (see Hill, 1997, p. 292). This does not only refer to opting-in or opting-out possibilities as provided in the provisions themselves, but also to agree on a set of arbitration rules or a choice of curial law that departs from the fall-back provisions of the law. Arbitration rules will usually provide for an arbitrator=s power to grant interim relief and thereby opt in section 39. This will trigger that the subsidiarity approach, especially the >effectiveness test= of section 44 (5) will become relevant. Access to the court is then restricted in line with this underlying concept. It should be noted that the court=s jurisdiction cannot be extended by parties= agreement or by choice of law. This means that arbitration rules or a choice of a law that provide for court access according to the free-choice model do not have any consequences insofar as they are inconsistent with section 44. The English courts are still bound to ground their jurisdiction on the matters listed in section 44 (2) and to follow the tests laid down in section 44 (3), (4) and (5). But choosing a curial law that prohibits court support in matters of interim relief (e.g. New York law) would be regarded as opting out of section 44. The same principle applies to section 42. It is not possible to bypass the hurdles set up in section 42 by agreeing on a law that provides for direct court assistance such as Swiss law. Agreeing on a different curial law only has consequences for the scope of power transferred to the arbitrator and the manner in which this is done. For example, if German law is stipulated in the arbitration agreement to govern interim relief, an arbitrator would have unlimited power to grant interim relief ex lege unless the parties agree otherwise. This would include the granting of a dinglichen Arrests, which is the functional equivalent of a Mareva injunction in German law. Thus, only with respect to the second issue a choice of law can be made. But it remains to be seen whether an English court would, for instance, enforce a German-style interim measure granted by the arbitrator under section 42 that is functionally equivalent to a Mareva injunction excluded under section 39. Generally speaking, once parties have opted in arbitrator-granted interim relief the subsidiarity model with its restrictions comes into full operation. It is not possible to bypass it while retaining the arbitrator=s power. Thus, care should be taken when drafting the arbitration agreement (compare Thomas, 1997, p. 409). 4.1.4 Summary The new English law has adopted an original approach with regard to interim measures of protection in arbitration. It established a system of court subsidiarity. The underlying philosophy regards the court as the last resort. To grant interim relief is in the first place allocated to the arbitrator. However, the parties have to opt in for it. The courts only step in under certain preconditions if the court-subsidiarity principle is applied. The preconditions for court-granted interim relief are high. But an ex parte Mareva injunction or an Anton Piller order are directly available from the court. Arbitrator-granted interim relief is enforced by the court. But hurdles are set up to keep interim measures as far as possible in the realm of arbitration. 4.2 Germany 4.2.1 Background Germany is, despite her economic strength, stability and modern infrastructure an unpopular arbitration venue. The new arbitration law was passed to change this situation. The hope is expressed that a modern and easily accessible legal framework will attract arbitration parties to Germany. Especially German-speaking East Europeans are a target group. The >German Institution for Arbitration= is a distinguished national arbitration centre. Its rules are currently being revised to suit the new law and to provide a state-of-the-art institutional framework for international commercial arbitrations. Germany now qualifies as a Model-Law country. The new arbitration law - in force since 1 January 1998 - incorporates the provisions of the Model Law almost literally into the Code of Civil Procedure (CCP). The German arbitration law has its place in the tenth book of that code since 1877. Germany is at the moment the only Model-Law country in Central Europe (Berger, 1998, p. 3). Neighbouring countries with an established arbitration tradition such as the Netherlands or Switzerland, having reformed their laws in the second half of the 1980s, did not adopt the Model Law. The Model Law became very popular with arbitration-law reformers in the 1990s. Law reform started officially in 1991. Prior to 1991 private initiatives prepared the ground in the arbitration community for a positive attitude towards the Model Law. In 1991 the Federal Ministry of Justice set up an expert commission to work out proposals for the reform of the German arbitration law. The commission was instructed to study the way in which optimum use could be made of the solutions provided in the Model Law. The report of the commission was published in 1994. It proposed the adoption of the Model Law with minor alterations. The report of the commission contains a detailed bill and a comprehensive commentary on each provision. The commentary formed the basis of the official travaux préparatoires of the governmental draft (Berger, 1998, p. 6). Thanks to the extensive travaux préparatoires, it will be easy for user and court alike to find out about underlying policies. This serves transparency. Adopting an internationally designed framework for arbitration was, in the absence of an elaborate and distinctly German arbitration culture, not a sacrifice of national legal culture. The old provisions, largely unchanged since 1877, were regarded as incomplete. Court decisions remedied the major inadequacies. The resulting piecemeal character of the law made access to it by foreigners difficult. The Model Law was regarded as the most suitable solution for international commercial arbitration. It is a thoughtfully prepared harmonizing legal regime providing not only for a maximum of party autonomy and a minimum of court intervention and mandatory provisions, but also easy access by foreigners. Berger (1998, p. 17) concludes, >today [.?.?.] counsel and arbitrators who know the Model Law also know the new German arbitration law=. 4.2.2 Survey of the solutions for the issues 4.2.2.1 The relationship between court-ordered interim relief and arbitration Underlying concept German law follows the free-choice approach of the Model Law. An arbitration party has the choice to apply either to the court or to the arbitral tribunal to obtain the interim measure of protection sought. There are no restrictions imposed on court access. There is no need for a party to seek permission from the arbitrator to apply to the court. Access to the court The relevant provision is section 1033 CCP. (Hereafter all provisions without further reference are those of the Code of Civil Procedure.) It restates article 9 of the Model Law, with the alteration that the court can only take interim measures of protection relating to the issue in dispute in the arbitration proceedings. This limits somewhat the scope of court support. Section 1033 codifies, firstly, that a party applying to the court for an interim measure of protection does not violate the arbitration agreement and, secondly, that the court can grant interim measures of protection. By stating the latter, section 1033 does not provide for the court=s jurisdiction to grant interim measures of protection; it only has a declaratory function. The court=s jurisdiction is to be established under provisions of the CCP dealing with competence for interim measures of protection. German law provides for two categories of interim measures of protection (Arreste and einstweilige Verfuegungen). Sections 916-945 prescribe their preconditions and deal with procedural matters regarding them. The first category comprises the dinglichen Arrest, which is functionally equivalent to a Mareva injunction. It is an in rem relief preventing the dissipation of assets. The same function has the persoenliche Arrest. The second category is constituted by the einstweiligen Verfuegungen. It is a broad term under which all sorts of remedies can be subsumed. It can be used to shape the relationship between the parties during adjudication and to preserve evidence. If the preconditions of an Arrest or an einstweiligen Verfuegung are fulfilled, the court is obliged to order it. It has only discretionary power with regard to the content of the order. In Germany the system of court-ordered interim relief is regarded as a special type of summary proceedings that is divided into two stages - a decision-finding stage and a separate decision-enforcement stage. Walker (1993, p. 61) writes: >The system of provisional remedies consists of two stages: during the first stage, the court=s decision takes place; it is determined whether there is a claim on securing the enforceability of a summarily determined subjective right. During the second stage, the enforcement of this decision takes place.= (my translation) An application for any interim measure of protection available under German law can be filed with a German court even if the seat of the arbitration is outside Germany or has not yet been determined (section 1025 (2)). The rules regarding international competence determine in that case the court=s jurisdiction. They are principally not different from the ones applied domestically according to the doctrine of double-functionality (see Schack, 1996, pp. 165-166). If assets of a party to a foreign arbitration are located in Germany, it will be possible to prevent their dissipation by means of interim relief applied for at a German court. Parties cannot opt out of court protection in matters of interim relief (LG Frankfurt). 4.2.2.2 The arbitrator=s competence to order interim measures of protection The arbitrator=s competence to order interim measures of protection is a novelty in German law. Section 1041 gives the relevant provisions. Subsection 1 adopts article 17 of the Model Law with minor changes in its wording. It spells out the arbitrator=s power to grant interim measures of protection with respect to the subject matter in dispute. The arbitrator can only act at the request of a party. Whether or not to grant an interim measure of protection is decided by the arbitrator on a discretionary basis. This is different with regard to court interim measures as mentioned. The judge has to order a measure if its preconditions are fulfilled. It has already been observed that this difference might lead to tactical decisions (Kronke, 1998, p. 264). To allow such a difference - constituting an advantage of court-ordered relief - shows that the policy of free choice is taken seriously. Subsection 1 empowers the arbitral tribunal to order that the applying party has to provide appropriate security in connection with the measure sought. The arbitral tribunal is not limited to German-style interim measures of protection. This can be inferred from interpreting section 1041 (2) (2). This provision states that the court may recast an arbitrator=s order if necessary for the purpose of enforcing the measure. The recast provision is in place to ensure that arbitrator-issued orders are translated into orders which fit the German enforcement system with its strict certainty principle (Bestimmtheitsgrundsatz) (see Bundesministerium der Justiz, 1994, p. 139). If German-style orders had to be granted by the arbitrator, there would hardly be a need for recasting. It arises if an arbitrator orders a foreign-style interim measure of protection such as, for instance, an English Mareva injunction. Can an arbitrator order ex parte interim relief under the new German law? Section 1042 (1) seems to bar obtaining such a relief from an arbitrator. The provision lays down the general rule of procedure that each party shall be given full opportunity to present its case. The wording of the rule indicates an infringement if an arbitrator acts on an ex parte basis. However, section 1042 (1) is a general rule of procedure. General rules provide for exceptions even without mentioning them expressly. An example of this relationship between rule and exception is the German constitutional guarantee of audiatur et altera pars, which corresponds to the general rule of section 1042 (1). Article 103 (1) of the German constitution enshrines this general rule without mentioning any ground for exceptions. It was argued before the Constitutional Court (the Bundesverfassungsgericht) that a judge=s order granted on an ex parte basis infringes art. 103 (1) of the German Constitution. The Constitutional Court held that - as an exception to the rule of art. 103 (1) of the German Constitution - an ex parte order does not infringe this constitutional provision if an ex parte procedure was necessary to secure interests that would otherwise be endangered. It is an exception to the rule born out of necessity to ensure the effectiveness of adjudication. A similar line of argument can be developed to provide for an exception with regard to section 1042 (1). Support for arbitrator-granted ex parte orders can be found in section 1063 (3) (1). This rule lays down that a judge can grant permission to enforce an interim measure of protection without prior hearing of the party opposing the application. The imposition of a mandatory hearing would have rendered any argument in favour of arbitrator-granted ex parte orders senseless. A hearing at the permission stage would prevent the necessary surprise effect, which is targeted at with ex parte relief. Thanks to section 1063 (3) (1), arbitrator-granted ex parte orders make sense. The legal framework for their efficiency is in place. 4.2.2.3 The enforcement of arbitrator-granted remedies Section 1041 (2), (3) and (4) as well as sections 1062 (1), (2) and 1063 (3) are the relevant provisions in this context. Section 1041 (2) states that a court may permit enforcement of an arbitrator-granted interim measure of protection unless application for a corresponding interim measure of protection has already been made to a court. The latter part of the provision ensures that two different courts will not be occupied with the same matter and thereby prevents contradicting court decisions (Schumacher, 1998, p. 11). As mentioned above, an application for court-ordered interim relief is made to the competent court as determined by general civil procedural law. The competent court will be one of first instance. Arbitration law (section 1062 (1) No. 3) provides for the competence of a Court of Appeal (Oberlandesgericht) with regard to granting permission to enforce an arbitrator=s interim measure. Generally, the Court of Appeal designated in the arbitrator=s order will be the competent court (section 1062 (1)). In Germany, each of the sixteen Federal States has at least one Court of Appeal. Practice will show whether arbitration applications will centre at one or two Courts of Appeal. This could foster specialization of counsel and bench. In Germany, counsels can only appear before one Court of Appeal. An arbitration specialist registered with the Court of Appeal in Cologne therefore cannot appear in Hamburg, Berlin or Frankfurt himself. He can, however, instruct local counsels. With section 1062 (2), German law provides a new solution for international commercial arbitration. The provision is rather hidden; its content is, however, very progressive as mentioned above. It reads: >If the place of arbitration in cases referred to subsection 1 [.?.?.] No. 3 [enforcement of an order for interim measures of protection by the arbitral tribunal] is not in Germany, the competence lies with the Court of Appeal where the party opposing the application has his place of business or habitual residence, or where assets of that party or the property in dispute or affected by the measure is located [.?.?.].= Section 1062 (2) gives competence to a court to enforce an arbitrator=s order even if the seat of arbitration is outside Germany. This provision renders arbitrator-issued interim relief enforceable in Germany if ordered at the seat of arbitration in Switzerland, France or elsewhere. It is a unilateral step to solve the cross-border problem with respect to Germany. The court exercises discretion with regard to granting permission to enforce an arbitrator-granted interim measure. The threshold of preconditions necessary for enforcement of arbitrator-granted orders is likely to be rather low. The Commission Report (Bundesministerium der Justiz, 1994, p. 139) refers to preconditions of whether there is a valid arbitration agreement and whether the order granted is not wholly misbalanced (Verhaeltnismaessigkeitskontrolle). This could be the case if an arbitrator=s order seriously hampers the exercise of a party=s business. Section 1041 (2) (2) provides the courts with the competence to recast an arbitrator=s order if necessary for the purpose of enforcing the measure. This will prevent courts refusing to grant permission because the arbitrator=s order does not fit the German system. As argued above, it enables an arbitrator to grant any remedy he likes. He need not stick to German-style relief with which he and the parties might not be familiar. Since, generally, (Germany is the new exception to the rule) arbitrator-granted interim relief is only enforceable by the courts of the country in which the arbitration seat is located, it would be advisable for an arbitrator to try to adjust to the style of interim relief common in the host jurisdiction, to ease enforcement. Recasting is not an easy task, as can be gathered from a look at the >translation= practice under the regime of the Brussels Convention (Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters, signed at Brussels on 27 September 1968). A translation of a Mareva injunction into German law under this regime by the Karlsruhe Court of Appeal serves to illustrate the difficulties (OLG Karlsruhe). The court discussed different ways of translating a Mareva injunction into German law, to meet the preconditions of the certainty principle (Bestimmtheitsgrundsatz). It proved to be more difficult than might have been suggested at first sight. A translation of a Mareva injunction into a dinglichen Arrest was ruled out, equally the transfer to an einstweilige Verfuegung. The court decided to enforce it under section 890 CCP, which bears an injunctive title. Zuckerman and Grunert (1996, p. 102) regard this option as less advantageous to the plaintiff than any other of the options ruled out by the court. However, the decision shows that it is too simple to assume that a Mareva injunction can automatically be translated into what is seen as its functional equivalent, the dingliche Arrest. Section 1041 (3) states that the court may, upon request, repeal or amend its decision to give leave to enforce the arbitrator=s order. The applicant can be either of the arbitration parties. The last subsection (section 1041 (4)) deals with damages which are payable by the party that applied for enforcement of an order which later turns out to have been unjustified from the outset. The claim for damages can be put forward in the pending arbitral proceedings. The subsection incorporates a general principle of interim relief. Interim relief is a risk-management mechanism. The party whose alleged right is in greater risk during the time that elapses until final adjudication can obtain an order. If the risk allocation turns out to be wrong, compensation must be paid, as the interim measure will have affected the opposing party. 4.2.3 On the restrictions of party=s choice as prescribed in the mandatory provisions of the lex arbitri German law now adheres to the territoriality principle (section 1025 (2)). This means that all German mandatory provisions must be observed if the seat of arbitration is in Germany. One of the mandatory rules is that the parties cannot opt out of court intervention with regard to interim relief; they can opt out of the arbitrator=s power to grant interim relief or the power of the court to enforce any order given by the arbitrator. For an arbitrator to grant a foreign-style interim measure of protection, no choice of law with respect to the interim remedies available needs to be made. Arbitration rules or a choice of curial law could provide for a restriction of court access by adopting a rule or a law that provides for an arbitrator=s permission to have access to the court. If this does not amount to an arbitrary ban on court access, German courts will likely recognize it. It could further be possible to limit the jurisdiction of an arbitrator by choosing a law that only allocates the granting of certain interim measures of protection to an arbitrator. German law is very compatible with arbitration rules and leaves wide room for a party=s choice. 4.2.4 Summary The German provisions on interim relief in arbitration provide for a free choice of a party to decide where to apply for the remedy sought. German law tries to make arbitrator-granted interim relief equally effective to court-ordered relief to provide for a real choice. It does not change the provisions of the Model Law but provides for a very progressive solution to the issue left open in the Model Law - the enforcement of arbitrator-granted interim relief. It remains to be seen how strictly the court will exercise control before it grants permission to enforce an arbitrator=s interim order. It is to be welcomed that an arbitrator can order interim relief of any style. The court will translate it so that it fits the German-style enforcement mechanism. There is reason to assume that an arbitrator can order interim relief on an ex parte basis and that courts will grant permission to enforce it on an ex parte basis. The original new solution German law provides concerns the enforcement of arbitrator-granted interim relief if the seat of arbitration is outside Germany. This is a unilateral step to provide for cross-border efficiency of arbitrator-granted interim relief. In the absence of an international regime, this solution could act as a role model for other countries. 4.3 Hong Kong 4.3.1 Background Hong Kong is an established Asian arbitration venue. The Hong Kong International Arbitration Centre (HKIAC) was created in 1985 to provide an independent forum for dispute resolution. Since its foundation, the number of disputes referred to the centre has increased steadily. By 1996 the centre had administered a total of 1144 arbitrations, the majority of which were construction and maritime disputes. But arbitrations taking place in Hong Kong are not limited to the HKIAC. Institutional arbitrations under the rules of all major international arbitration institutions and ad hoc arbitrations are conducted in Hong Kong. The transfer of sovereignty over the territory to the People=s Republic of China on 1 July 1997 has, however, entailed a setback for Hong Kong as the prime venue for arbitration with parties from the People=s Republic of China. This is due to the uncertainty of the application of the New York Convention between Hong Kong as Special Administrative Region (SAR) and the People=s Republic of China. The legal framework for international commercial arbitration is the Model Law. It was adopted in 1990. Section 34C (1) of part IIA (International Arbitration) of the Hong Kong Arbitration Ordinance (Chapter 341 of the Laws of Hong Kong) renders chapters I to VII of the Model Law applicable for international arbitration. An important amendment came into force on 27 June 1997, changing the structure of the Arbitration Ordinance. The provisions for international commercial arbitration are now not only found in part IIA but also in part IA. This part contains the >Provisions Applicable to Domestic and International Arbitrations=. Among them are provisions that deal with interim measures of protection. Unless being subject to a section of part IA, the provisions of the Model Law apply. Articles 9 and 17 Model Law are not subject to any section of part IA. The sections of part IA dealing with interim measures (sections 2GB, 2GC and 2GG) thus only supplement and do not supersede the Model Law provisions. Section 34E, which dealt with >powers of Court in relation to evidence and interim relief=, ceased to have force. Section 2GC now deals with these matters. 4.3.2 Survey of the solutions for the issues 4.3.2.1 The relationship between court-ordered interim relief and arbitration Underlying concept Hong Kong law follows the free-choice approach of the Model Law. However, the new amendment lays down a conflict or limitation rule for the case in which an application is concurrently made to court and arbitral tribunal. It grants discretion to the courts to decline its jurisdiction under certain circumstances. This limits the free access of arbitration parties to the court. The court will determine the extent of the limitation. There are no limitations prescribed in the law for granting leave to enforce the arbitrator=s order. But it was already assumed that >the courts are likely to require application to be made first to the tribunal for an appropriate sanction= (Morgan, 1997, p. 81) before the courts grant leave to enforce an arbitrator=s order. If the courts follow the line Morgan laid out, Hong Kong would ultimately align itself with a subsidiarity approach. Until now no cases are reported on the interpretation of the relevant provisions. Access to the court The first issue is regulated by article 9 Model Law and section 2GC. Article 9 Model Law >codifies the dual principle that, first, a party does not waive its right to go to arbitration by requesting (or obtaining) interim measures of protection from a national court, and, second, that a national court is not prevented from granting such measures by the existence of an arbitration agreement= (Holtzmann and Neuhaus, 1989, p. 332). Section 2 GC is about >Special powers of Court in relation to arbitration proceedings=. The section is not restricted to interim measures of protection. It rather deals with different matters that are at stake during an arbitration and grants the court power in respect of them. Relevant for this context is the court power to grant an interim injunction or direct any other interim measure to be taken (section 2GC (1) (c)). Interim measures available from court are similar to English measures. They include, for instance, Mareva injunctions, Anton Piller orders and interlocutory injunctions. Section 2 (1) defines >court= as the High Court. With the transfer of sovereignty to the People=s Republic of China on 1 July 1997, the High Court became the Court of First Instance (article 81 Basic Law of the Hong Kong Special Administrative Region of the People=s Republic of China). Section 2GC (5) states that the powers conferred can be exercised by the court irrespective of whether or not similar powers may be exercised under section 2GB by the arbitrator in relation to the same dispute. This clarifies the fact that the courts= power is not restricted by a concurrent arbitrator=s power. This is even true for the case in which an arbitrator exercises his concurrent power. Subsection 6 provides a guideline for this case. It is for the court to decide whether the tribunal is the more appropriate forum to grant the interim measure of protection. It can be assumed that it is appropriate for a judge to decline to make an order if the application has not been made in good faith or is a delaying tactic (see Morgan, 1997, p. 58). The guideline can have great relevance as the arbitrator has nearly all the powers the court has. Subsection 6 provides for a conflict rule to guide the court. It is not a conflict rule as encountered in the Arbitration Act 1996 as the court has discretion. The English rule excludes court power under certain circumstances ex lege. Hong Kong Law has therefore not departed from the free-choice approach incorporated in article 9 Model Law by introducing section 2GC. The courts could, however, cause such a departure with their interpretation. The parties cannot opt out of court support (Morgan, 1997, p. 55). This is a setback for party autonomy. The exhaustive list of court powers prevents dispute over whether a certain interim measure is obtainable from the court. Security for costs is only obtainable from the arbitrator. 4.3.2.2 The arbitrator=s competence to order interim measures of protection Article 17 Model Law and section 2GB are the two relevant provisions with respect to this issue. Article 17 Model Law >provides that, unless otherwise specified by the parties, the arbitral tribunal has the power to order interim measures of protection in respect of the subject matter of the dispute= (Holtzmann and Neuhaus, 1989, p. 530). It further states that the arbitral tribunal may require any party to provide appropriate security in connection with such measure. Section 2GB spells out the general powers exercisable by the arbitral tribunal. It thus adopts a similar structure as section 38 Arbitration Act 1996. Power under section 39 Arbitration Act 1996 is incorporated in section 2GB. Section 2GB is not restricted to interim measures of protection. An arbitrator has power to order security for costs under section 2GB (1) (a) and to grant interim injunctions or directing other interim measures to be taken under section 2GB (1) (f). These are the two powers relevant to interim measures of protection. The arbitrator=s competence with regard to interim injunctions and other interim measures is spelled out very roughly. No examples are given. It can be assumed that the power to grant interim measures of protection encompasses all sorts of interim relief. There are no restrictions other than those inherent in the arbitrator=s jurisdiction. His orders cannot bind third parties (Morgan, 1997, p. 49). Therefore ex parte Mareva injunctions and Anton Piller orders are principally available from the arbitrator. But they do not have third-party effect. It is to be welcomed that the Arbitration Ordinance provides the arbitrator=s power ex lege. The parties cannot opt out of the arbitrator=s power to grant interim measures of protection (argumentum e section 2GB (9)). Since the parties can always apply concurrently to the courts and it is for the court to decide whether it will grant the remedy sought, no harm is done by excluding an opt-out possibility. The lack is, however, hard to reconcile with the principle of party autonomy. 4.3.2.3 The enforcement of arbitrator-granted remedies Section 2GG provides for an enforcement mechanism of decisions of the arbitral tribunal in a summary procedure. It refers to all sorts of arbitral decisions, including awards, orders and directions. Section 2GB (1), which spells out the arbitrator=s competence, refers to orders and directions. Therefore it can be assumed that an arbitrator-granted interim measure of protection comes in the form of an order or direction and not in the form of an interim award. With leave of the court, arbitrator=s orders are enforceable in the same way as an order of the court. This means that the arbitrator=s order carries by leave of the court the contempt of court threat. This threat will also reach third parties that aid in breaking the order. Directions from an arbitrator with respect to interim measures will be enforced as court directions. Interim measures other than interim injunctions are granted by the arbitrator in the form of a direction. No cases have as yet been reported which interpret section 2GG. It can be assumed that the court must be satisfied that there is a valid arbitration agreement and that the arbitrator had power under section 2GB. Section 2GG does not put up any further preconditions that must be met before the court can enforce an arbitrator=s order or direction. Morgan (1997, p. 81) suggests that a >party under the Hong Kong provisions would not be justified in seeking to enforce an order or direction of the tribunal immediately following the first instance of default by the other party=. He adds: >The courts are likely to require application to be made first to the tribunal for an appropriate sanction.= If the courts follow these suggestions, the Hong Kong solution would be aligned with the one the Arbitration Act 1996 provides. As section 2GG is almost identical to section 12 (5) Singapore International Arbitration Act, Singapore cases - not yet reported either B on the Singapore provision can have persuasive authority for the interpretation of this section and vice versa. 4.3.3 On the restrictions of party=s choice as prescribed in the mandatory provisions of the lex arbitri Section 34C (1) in connection with article 1 (2) Model Law lays down the territoriality principle. Thus all Hong Kong mandatory provisions have to be observed. Among the mandatory provisions count the arbitrator=s power to grant interim relief (Morgan, 1997, p. 45). For instance, if the parties agree on Italian law as the curial law, which would govern the question of interim relief, the following consequence would be triggered. Italian law (article 818 Code of Civil Procedure) states that only courts can grant interim relief. This Italian provision conflicts with the mandatory provision of section 2 GB (1) (f). An arbitrator could therefore grant interim relief despite the exclusion under Italian law. The power of the court to order interim relief can neither be ousted by the arbitration agreement. To choose a foreign curial law with respect to interim measures of protection does not make much sense in Hong Kong. All issues are covered by provisions that prevail. This also sets limits to the parties= choice of arbitration rules which depart from the mandatory provisions. 4.3.4 Summary Hong Kong has adopted an interesting legislative technique. It did not supersede the provisions of the Model Law on interim relief with new provisions but supplemented them with provisions common to domestic and international arbitration. The solutions provided for in the amendment do not adopt the subsidiarity model of the English Arbitration Act 1996 but constitute a distinct approach in line with the free-choice model. The courts are entrusted with the power to decline jurisdiction under certain circumstances. It remains to be seen whether, by following English case law, the interpretation of the Hong Kong provisions will align with the underlying policy of the English Act rather than providing a distinct model. 5. Comparative aspects The comparative part of this article will focus on the different legislative techniques used and their consequences for both the workability of the different new legislation and the realization of party autonomy in practice. It will address the three main issues. The restrictions of party=s choice as prescribed in the mandatory provisions of the lex arbitri will be evaluated. 5.1 The relationship between court-ordered interim relief and arbitration 5.1.1 Underlying concepts compared The philosophy underlying the English approach - to limit court intervention as far as possible - serves the need for arbitration. Court applications can seriously delay and hamper the private dispute settlement mechanism. It seems ideal to incorporate interim measures of protection into the realm of arbitration. But court intervention becomes necessary at the enforcement stage. Interim measures of protection are a hybrid species. The role of the courts cannot be excluded. English legislation takes this into account when it provides for a special enforcement mechanism for arbitrator-granted interim relief. But still, arbitrator-granted interim relief is not equally effective as court-ordered interim relief. This is conceded in the new Act when effectiveness was chosen as the predominant rule to decide whether courts have jurisdiction or not. The English Act confers upon the court the assessment of whether or not the arbitrator can order interim measures of protection in an equally effective manner. The Act does not entrust the assessment to the parties. It would be reasonable to assume that a party can best assess what is the most efficient way to obtain the remedy sought. By adopting the new approach, the courts= influence is shifted, it is not cut back. The sophisticated mechanism of >effectiveness test= and >urgency test= could further stimulate dispute over the interpretation and therefore be a source of delay and frustration. If this is the case, the goal of limiting court intervention to smoothen the arbitral process is missed. With regard to the Hong Kong approach, the key lies with the judges. They will interpret the new provisions and can give effect to the laws in accordance with either of the models. They should monitor how the subsidiarity model works in practice before aligning with it. An advantage of the Hong Kong Model is that it gives the judges discretion as to whether to decline to exercise their jurisdiction or not, if, on the facts of a given case, it may be more appropriate for the matter to be dealt with by the arbitration tribunal. Unfortunately, this discretion is limited to the situation where the arbitrator is concurrently dealing with the matter. This would only occur in rare cases. It is to be welcomed that the lex arbitri entrusts the courts with such a discretion as it can support arbitration by giving a ground to rule out dilatory and tactical applications. It would have been preferable if the Hong Kong legislator had dropped the connection between appropriateness and concurrent arbitral proceedings that limits the exercise of this discretion to rather rare cases as mentioned. It would have been equally appropriate for German law to include such a discretionary power to act as a deterrent against tactical applications. As the law stands now, a German court has to grant the interim remedy if its preconditions are met, thus the application has to be considered on its merits and cannot be rejected on a preliminary ground as provided under Hong Kong law in certain circumstances. The free-choice Model has advantages as it leaves the assessment of where to apply to obtain the most efficient provisional remedy to the parties. It minimizes grounds for dispute over the dispute resolution mechanism with its simple and clear mechanism. In the German and Hong Kong versions, it provides for the effective enforcement of arbitrator-granted interim relief and, thereby, gives parties a real choice. 5.1.2 Access to the court In all three countries arbitration parties= access to the court is provided for. However, unlimited access is only provided under German and Hong Kong law. Under English law, court access it limited in line with the court-subsidiarity model. Although England and Hong Kong follow different models, they have a common legislative technique that differentiates them from Germany. English and Hong Kong arbitration law confers only a catalogue of specific powers upon the court. The transparency of this legislative technique helps to exclude specific interim measures from the court=s power. This facilitates opting-out (as possible under English law) but also helps to circumscribe what remedies the courts can provide. Neither a court in England nor a court in Hong Kong can order security for costs as this interim remedy is not found in the catalogue of powers on which the court can base its jurisdiction. These are advantages compared to article 9 Model Law as adopted in Germany, which generally refers to interim measures of protection without conferring actual power to the court. It is declaratory. Article 9 Model law includes all remedies regarded under domestic law (the lex fori of the arbitration seat) as interim ones. As they are likely to be different from the ones an adviser of a party is used to in his own jurisdiction, it becomes necessary for him to research the grounds establishing jurisdiction for a certain interim measure of protection and the remedies available under the law of the country where the seat is located to oversee the risks their availability might involve. This is difficult. It is easier to exclude certain matters and with them all remedies which serve this end. This helps to incorporate solutions in an arbitration agreement without the need of having a detailed knowledge of the procedural law of the seat=s jurisdiction. This is, however, necessary if one wants to exclude a certain type of interim remedy such as the référé procedure in France or an Anton Piller order in England. But as already has been mentioned, any exclusion of interim measures of protection either by remedy or by matter (on which a remedy can be based) carries the risk of excluding the remedy the own client may need at a later stage. If a legislator allows parties to opt out of court-ordered interim relief, the English drafting technique is superior to the one of the Model Law as it provides for more transparency. Besides the transparency advantage, the English or Hong Kong legislative technique seems generally to be an especially advantageous solution for common-law jurisdictions. In England and Hong Kong, as examples of common-law jurisdictions, interim relief is arranged in remedies. Different remedies such as interim payments, the appointment of a receiver, Mareva injunctions and security for costs form the group of diverse remedies that are identified as interim relief. In Germany, as an example of a civil-law country, only two categories of interim relief are available, as has been mentioned. Under these two categories all interim measures of protection can be subsumed. The absence of remedy-style interim relief under German civil procedural law matches the Model-Law solution with its general reference to interim measures of protection. It is an adequate solution for German law. It is, however, clouding the view if interim relief comes in a Hong Kong or English-style remedy system. A catalogue of remedies does not conflict with the Model Law. The new Hong Kong provision listing the court=s powers with regard to interim measures of protection supplements article 9 Model Law. Indeed, it specifies what the Model Law calls interim measures of protection for Hong Kong. The difference in legislative technique thus coincides with the difference in legal family. 5.2 The arbitrator=s competence to order interim measures of protection All three laws provide for the arbitrator=s competence to order interim relief. Under German and Hong Kong law, no restrictions are imposed on the arbitrator=s power. English law excludes certain remedies such as ex parte Mareva injunctions and Anton Piller orders from an arbitrator=s competence. But these two are remedies an arbitration party would most probably not apply for with an arbitrator. Again, a difference can be identified with regard to the drafting technique. Germany and Hong Kong follow the same approach. The English lex arbitri departs. Whereas German and Hong Kong law provides for an arbitrator=s power to grant interim measures of protection ex lege unless the parties agree otherwise, parties have to confer such a power to the arbitrator under English law. This makes a fundamental difference. The English opt-in version was praised as an >elegant application of the fundamental concept of party autonomy. Not only are parties free to agree that a tribunal will have the power to grant interim injunctions but such a power will not be vested in a tribunal unless the parties agree to it.= (Thomas, 1997, p. 405). This is true. But there is reason to rethink whether party autonomy is indeed realized if an opt-in version is chosen in the lex arbitri. It might turn out to be a trap for the unwary in ad hoc arbitrations - a trap in the sense that the English drafting technique departs from the commonly used one. The Model Law and, in line with it, generally most modern legislation provide for the arbitrator=s competence ex lege unless the parties have agreed otherwise. Ad hoc arbitrations are primarily shaped by the tailor-made arbitration clause. In the absence of an institutional framework they depend strongly on the legal framework of the country in which the arbitration has its seat, to fill gaps in the clause. One of the gaps in the clause is very likely to be the competence of the arbitrator to grant interim measures of protection (see Blessing, 1997, p. 5 footnote 4). Only if the legislator takes the drafting practice (and inadequacies) into account, the solution will be suitable. This is a conclusion valid for a free-choice model. The aim is to give the parties a choice between arbitrator-granted and court-ordered interim relief. To achieve this end, a legislative framework must be in place that even takes drafting inadequacies into account. But the English system adheres to the court-subsidiarity approach. To opt in arbitrator-granted relief triggers limited access to the courts under this concept. Whether parties appreciate this consequence can be doubted. Indeed it might be regarded as the bigger trap. Under the new English law the decision whether to opt in section 39, has to be taken consciously as consequences flow from it that are outside the parties= power to control by agreement or choice of law. Therefore the opt-in version can be regarded as the more suitable realization of party autonomy in the context of a court-subsidiarity model. But this conclusion raises further doubt about the general suitability of the court-subsidiarity model for arbitration. Its need for conscious decision taking on a minor question such as interim relief does not adequately consider the drafting practice. Arbitration clauses are often drafted in a rush without much attention to details of the lex arbitri. The different legislative techniques in conferring power are also encountered with respect to granting power to the arbitrator. In line with the Model Law, German law confers general power to the arbitrator to order interim measures of protection. This includes German-style and foreign-type interim measures of protection. The power granted is therefore very wide. Hong Kong law spells out the powers very widely as well, but again refers to certain remedies. England is more specific than Hong Kong. This is due to the differentiation in English law between power granted to the arbitrator ex lege for certain interim measures such as security for costs under section 38 and the power that section 39 provides in case the parties opt in for it. The English and Hong Kong legislative technique serves transparency at the stage of drafting an arbitration clause but might lead an arbitrator to solely focus on English-style interim remedies. With regard to the arbitrator=s competence to order interim relief the differences in legislative techniques have their cause not only in the difference of legal family but also in the model adopted. 5.3 The enforcement of arbitrator-granted interim relief The enforcement issue is left open in the Model Law. Generally speaking, two different concepts of enforcement have emerged in modern legislation (compare Sanders, 1996, p. 115). Some legislators count provisional orders by arbitrators as awards. They are enforceable according to the regular award enforcement procedures including the possibility of a setting-aside application. The process is a rather lengthy one. The Dutch legislator has adopted this approach in 1986. The relevant provisions in the Dutch Code of Civil Procedure read: >The arbitral tribunal may render an interim award= (article 1049) and >A decision rendered in summary arbitral proceedings shall be regarded as an arbitral award [.?.?.]= (article 1051(3)). The former refers to normal arbitrator-granted interim relief and the latter to the arbitraal kort geding which is a specialty of Dutch arbitration law shifting the kort geding procedure that is similar to the French référé procedure into the realm of arbitration. The parties have to opt in for it. Swiss law, on the other hand, in 1989 adopted the approach to provide a special mechanism for the enforcement of arbitrator-granted interim relief in article 183 (2) Private International Law Statute. The local judge can be asked to assist the tribunal in enforcing the order. The judge applies his own law. The order is not regarded as an award and, consequently, the provisions on the setting-aside procedure do not find application. The judge only controls whether there is a valid arbitration agreement, whether the arbitral tribunal has been established validly and whether the remedy granted by the arbitrator does not conflict with the Swiss ordre public (see Berti, 1995, p. 1540). The English, German and Hong Kong solution for enforcement can be grouped under the Swiss approach. Thomas (1997, p. 407) expresses the hope that arbitrator=s orders granted under section 39 could also be enforced as an award under section 66. This is, however, doubtful as he concedes himself (supra). As mentioned above, the enforcement of an arbitrator=s order as an award is a lengthy process in the domestic context. In that context, the Swiss approach is more suitable. However, it does not provide an answer for cross-border enforcement. Court support is restricted to Switzerland. The award solution could have advantages in the cross-border context if the awards would be enforceable under the New York Convention. This is, however, not the case (supra). Therefore the >award track= enforcement mechanism has no advantage. The English, German and Hong Kong legislators have thus decided on the better alternative when following the Swiss model. The English solution is, however, more elaborate than the simple Swiss approach as it includes a mechanism restricting court support to a course of last resort. German and Hong Kong law does not impose any restrictions on court support, and therefore follows the plain Swiss approach. The German lex arbitri provides for recasting of foreign-type measures of protection to meet the preconditions of the German enforcement mechanism. Only German law provides for a provision that the party that enforced an arbitrator-granted remedy has to pay damages if a measure proves to have been unjustified from the outset. Neither English nor Hong Kong law deals with the question of an undertaking as to damages from a party applying for court enforcement of an arbitrator=s order. It would be appropriate to demand such an undertaking. Only German law provides for the enforcement of arbitrator-granted interim relief from an arbitral tribunal having its seat outside Germany. It thereby provides for a unilateral solution for the cross-border enforcement of arbitrator-granted interim relief. All three countries try to make arbitrator-granted interim relief as efficient as possible by providing for an enforcement mechanism. However, all these efforts cannot change the fact that arbitrator-granted interim relief will always be a step behind court-ordered relief. This is true for all countries. The process of granting interim relief and its enforcement can broadly be described in procedural steps. In German law, the interim measure is ordered and can generally be enforced straightaway because there is no need for a Vollstreckungsklausel (section 929 subsection 1 Code of Civil Procedure), which is an administrative confirmation of the enforceability. Thus two steps are needed to enforce a court-ordered interim measure. If the arbitrator orders the same measure, three steps are needed. The additional step of court permission could prove to be one step too much if urgent relief is required. In England an arbitrator can order injunctive relief and the courts step in to provide for enforcement thereof. In Hong Kong arbitrators are equally empowered to order interim measures of protection unless the parties agree otherwise. The arbitrator=s interim orders are enforceable by leave of the court in the same manner as if they were orders made by a court. Thus in England and Hong Kong two steps are required for arbitrator-granted interim relief to have more than contractual effect between the arbitration parties. That is one step less than in Germany but court orders in these countries have immediate effect and are thus a very efficient one step mechanism. There is no special enforcement procedure necessary like in Germany. The order is fortified by contempt of court, a notion unknown to German law. Thus arbitrator-granted interim relief always calls for an additional step in all three countries if enforcement is at stake. This causes delay. Therefore parties might prefer to apply to the courts to get the kind of relief they perceive as more efficient. Even if the enforcement limitation of the arbitrator=s jurisdiction is overcome by court intervention, there still remains the question of efficiency. 5.4 On the restrictions of party=s choice as prescribed in the mandatory provisions of the lex arbitri England is the only country among the ones compared that does not provide for any mandatory provision with regard to interim measures of protection. German law prescribes that parties cannot opt out of interim measures granted by the court. Hong Kong law lays down that the parties cannot opt out of the arbitrator=s power. Nor is there any possibility that they could do so with regard to the respective court powers. At first glance it seems that party autonomy is given most room in England. But a second glance will reveal that indeed the English lex arbitri restricts the parties most, despite the many possibilities to opt in or out of the fallback provisions of the Arbitration Act 1996. The parties are free to opt in for arbitrator-granted interim relief. However, if they do so, they come under the court-subsidiarity regime. There is no possibility to bypass the limitations of court power by agreeing on arbitration rules or a foreign curial law. Once the choice for arbitrator-granted relief is made, the parties have to play according to the rules of the lex arbitri if they want provisional relief from the court. Furthermore, not only direct access to the court is limited but also court support with regard to enforcing arbitrator-granted interim relief. Parties cannot change this jurisdictional limitation by agreement. Parties= only real choice is to decide whether they want to opt in section 39 or not. If they agree on arbitration rules that provide for the arbitrator=s power to grant interim relief, they are stuck in the court-subsidiarity regime. There is no way out. Parties should therefore very well consider whether to opt in this provision. The consequences are beyond their control. The only good news is that ex parte Mareva injunctions and Anton Piller orders are available from the court without any limitations even if the court-subsidiarity regime applies. Under the German lex arbitri parties have a wide choice as long as they do not rule out the court=s power to grant interim measures. It is, for instance, possible to depart from the free-choice model by agreeing on a court-subsidiarity approach. The ICC Rules allow only recourse to the courts in >appropriate circumstances= (article 23 (2)). This is a limitation on court access reconcilable with German law. The arbitration law already grants the arbitrator the widest competence possible to decide on interim measures of protection. Parties can limit it. The German approach is thus very supportive with regard to parties= choices. Hong Kong law counts the greatest number of mandatory provisions. Parties= choices are limited. They cannot exclude certain interim measures of protection from the arbitrator=s competence. This is a delicate matter. The arbitrator is granted the utmost power possible. He can - theoretically - even grant a Mareva injunction or an Anton Piller order. With respect to security for costs, it is even only the arbitrator who can order them. But as long as the courts are unlimitedly open to arbitration parties, the arbitrator=s power will not force them to apply to the arbitrator. For instance, a Mareva injunction would in any case be more efficiently granted by the courts than by an arbitrator. There is no need for the additional step of court leave. Taking into account the additional step any restriction of court access can mean harm to the parties. Hong Kong courts should take this into account when interpreting the provisions which might result in a limitation of court jurisdiction. 6. Conclusion Interim measures of protection in arbitration cannot get around court involvement if effectiveness through the threat of enforcement is to be ensured. The critical question is therefore how best to shape the interface between arbitration and the courts. The new legislation in each country has provided for new answers to old questions. The English answer is an original one that departs from any common concept. It offers a sophisticated and systematically stringent system of court subsidiarity. The underlying aim is to reduce court interference by allocating only limited powers to the court. The aim is principally to be welcomed. The system itself is a fine example of an elaborate court-subsidiarity model. But from a wider perspective any court-subsidiarity model shows inadequacies. As outlined, any concept of court subsidiarity needs a conflict rule that determines the limits of court power. The interpretation of the conflict rule is allocated to the courts. It might give rise to dispute over its interpretation and does not reduce court influence; it rather shifts it. But cutting court interference is only necessary where court involvement has adverse effects on arbitration. If court involvement supports arbitration it is to be welcomed. With regard to interim measures of protection the interface between arbitration and the courts will ensure that parties are given the widest possible choice to decide - at the moment when a certain interim measure of protection is needed - on the most efficient track leading to it. To realize party choice, a legislative framework is needed that grants the arbitrator the widest power possible with respect to interim measures of protection ex lege and that provides for the enforcement of an arbitrator-granted order. When such a legal infrastructure is in place, parties will have a real choice. Further, there should be no mandatory provisions with regard to interim relief. If parties want to shape the interface differently by agreeing to arbitration rules or a curial law that departs from the fallback provisions of the lex arbitri of the arbitration seat, they should be free to do so. National law should as far as possible support the enforcement of arbitrator-granted relief even if the seat of the arbitration is outside the enforcement country. German and Hong Kong law, which follow the Model Law, provide for new solutions which should be taken into consideration by any future legislator. They will have to be proven in practice. But so far they represent the latest state-of-the-art solutions under the free-choice model. The English law will be a role model for any legislator considering the adoption of a court-subsidiarity model. It is, however, advisable to monitor its development in practice before following it. It remains to be seen whether English law in action will remain as distinct from German or Hong Kong law as their contrary underlying concepts suggest (or vice versa). Bibliography Andrews, N. (1994) Principles of Civil Procedure (London: Sweet & Maxwell). Berger, K.P. (1993) International Economic Arbitration (Deventer/Boston: Kluwer). Berger, K.P. (1998) >The Implementation of the UNCITRAL Model Law in Germany=, Mealey=s International Arbitration Report, 13 (1), 38-54. Berti, S.V. (1995) >Kommentierung des Artikels 183 IPRG=, in H. Honsell, N.P. Vogt and A.K. Schnyder (eds.), Kommentar zum Schweizerischen Privatrecht: Internationales Privatrecht (Basel/Frankfurt: Helbing & Lichtenhahn). Black, H.C. (1891) A Dictionary of Law (St. Paul: West). Blessing, M. (1997) >The Trouble with Interim Relief in International Arbitration=, (conference paper) International Arbitration Day: Global Perspectives, AAA/New York Bar Association/IBA, New York, 25 September 1997. Boesch, A. (1989) Einstweiliger Rechtsschutz in der internationalen Handelsschiedsgerichtsbarkeit (Frankfurt/Bern/New York/Paris: Peter Lang). Born, G.B. (1994) International Commercial Arbitration in the United States (Deventer/Boston: Kluwer). Bundesministerium der Justiz (ed.) (1994) Kommission zur Neuordnung des Schiedsverfahrensrechts: Bericht mit einem Diskussionsentwurf zur Neufassung des Zehnten Buches der ZPO (Bonn: Bundesministerium der Justiz). Collins, L. (1994) Essays in International Litigation and the Conflict of Laws (Oxford: Clarendon Press), pp. 1-16. Collins, L. (ed.) (1997) Dicey and Morris: The Conflict of Laws, Fourth Cumulative Supplement to the 12th edition (London: Sweet & Maxwell). Craig, W.L., W.W. Park and J. Paulsson (1990) International Chamber of Commerce Arbitration, 2nd edn. (New York/London/Rome/Paris: Oceana/ICC). Harris, B., R. Planterose and J. Tecks (1996) The Arbitration Act 1996: A Commentary (Oxford: Blackwell Science). Hill, J. (1997) >Some Private International Law Aspects of the Arbitration Act 1996=, International and Comparative Law Quarterly, 46 (2), 274-308. Holtzmann, H.M. and J.E. Neuhaus (1989) A Guide to the UNCITRAL Model Law on International Commercial Arbitration: Legislative History and Commentary (Deventer/Boston/The Hague: Kluwer). Knoepfler, F. (1997) >Les mesures provisoires et l=arbitrage international=, in A. Kellerhals (ed.) Schiedsgerichtsbarkeit (Zuerich: Schulthess), pp. 307-329. Knoepfler, F. and Ph. Schweizer (1984) >Les mesures provisoires et l=arbitrage=, in C. Reymond and E. Bucher (eds.), Recueil de travaux suisses (Zuerich: Schulthess), pp. 221-228. Kronke, H. (1998) >Internationale Schiedsverfahren nach der Reform=, Recht der Internationalen Wirtschaft (RIW), 44 (4), 257-265. Merkin, R. (1996) Arbitration Act 1996: An Annotated Guide (London/New York/Hong Kong: LLP). Miller, F. (1996) >Arbitration Act 1996: The Proceedings in Parliament, Westminster=, Arbitration, 62 (4), 250-257. Morgan, R. (1997) The Arbitration Ordinance of Hong Kong: A Commentary, 1997 Supplement (Singapore/Hong Kong/Malaysia: Butterworths). Rutherford, M. and J.H.M. Sims (1996) Arbitration Act 1996: A Practical Guide (London: FT). Sanders, P. (1996) >Arbitration=, Chapter 12 of M. Cappelletti (ed.), >Civil Procedure=, Vol. XVI of the International Encyclopedia of Comparative Law (Tuebingen/Dordrecht/Boston/Lancaster: J.C.B. Mohr (Paul Siebeck)/Martinus Nijhoff). Schack, H. (1996) Internationales Zivilverfahrensrecht, 2nd edn. (Muenchen: C.H. Beck). Schlosser, P. (1989) Das Recht der Privaten Internationalen Schiedsgerichtsbarkeit, 2nd edn. (Tuebingen: J.C.B. Mohr (Paul Siebeck). Schumacher, K. (1998) >Das neue 10. Buch der Zivilprozessordnung im Vergleich zum UNCITRAL-Modellgesetz ueber die internationale Handelsschiedsgerichtsbarkeit=, Betriebs-Berater (BB), Beilage 2 vom 19. 3. 1998, 6-17. Schwartz, E.A. (1993) >The Practice and Experience of the ICC Court=, in Conservatory and Provisional Measures in International Arbitration (Paris: ICC). Sutton, D.St.J., J. Kendall and J. Gill (1997) Russell on Arbitration, 21st edn. (London: Sweet & Maxwell). Thomas, D.B. (1997) >Interim Relief Pursuant to Institutional Rules under the English Arbitration Act 1996=, Arbitration International, 13 (4), 405-409. Van den Berg, A.J. (1981) The New York Arbitration Convention of 1958 (Deventer/Antwerp/Boston/London/Frankfurt: Kluwer). Walker, W.-D. (1993) Der einstweilige Rechtsschutz im Zivilprozess und im arbeitsgerichtlichen Verfahren (Tuebingen: J.C.B. Mohr (Paul Siebeck). Zuckerman, A.A.S. and J. Grunert (1996) >Anmerkung zu OLG Karlsruhe=, Zeitschrift fuer Zivilprozess International (ZZPInt.) 1, 96-102. Documents 1. Conventions New York Convention of 10 June 1958 on the Recognition and Enforcement of Foreign Arbitral Awards, 330 United Nations Treaty Series, p. 38, No. 4739 (1959), <http://itl.irv.uit.no/trade_law/doc/UN.Arbitration.Recognition.and.Enforcement.Convention.New.York.1958.html>. Brussels Convention on Jurisdiction and the Enforcement of Judgments in Civil and Commercial Matters of 27 September 1968, Official Journal of the European Communities, C189, 28 July 1990, p. 1. 2. Arbitration rules International Chamber of Commerce (ICC), ICC Rules of Arbitration (1 January 1989-31 December 1997) (Paris: ICC). International Chamber of Commerce (ICC), ICC Pre-Arbitral Referee Procedure (1990), <http://www.iccwbo.org/arb/36.htm>. International Chamber of Commerce (ICC), ICC Rules of Arbitration (1 January 1998), <http://www.iccwbo.org/commissions/Arbitration/rulesenglish.htm>. Singapore International Arbitration Centre, Arbitration Rules of the Singapore International Arbitration Centre (22 October 1997), <http://siac.tdb.gov.sg/rules.html>. UNCITRAL Arbitration Rules for International Commercial Arbitration, U.N. Doc. A/31/17 (1976). <http://itl.irv.uit.no/trade_law/doc/UN.Arbitration.Rules.1976.html>. World Intellectual Property Organization (WIPO) Emergency Relief Rules, WIPO Doc. ARB/AC/III/96/3 of 15 November 1996. 3. Other documents UNCITRAL Model Law on International Commercial Arbitration, U.N. Doc. A/40/17 (Annex I) (1985), <http://itl.irv.uit.no/trade_law/doc/UN.Arbitration.Model.Law.1985.html>. Departmental Advisory Committee (DAC), Report on the Arbitration Bill, February 1996, in Merkin (1996, p. 163). Practice Note (1997) 1 All ER 39. Slaughter and May (July 1997) The Arbitration Act 1996: Memorandum (London: Slaughter and May). Legislation England Arbitration Act 1996, Chapter 23. <http://www.hmso.gov.uk/acts/acts1996/1996023.htm>. Rules of the Supreme Court (Amendment) 1996, SI 1996/3219, in Sutton, Kendall and Gill (1997, p. 548). The Arbitration Act 1996 (Commencement No. 1) Order 1996, SI 1996/3146, in Sutton, Kendall and Gill (1997, p. 544). The High Court and County Courts (Allocation of Arbitration Proceedings) Order 1996, SI 1996/3215, in Sutton, Kendall and Gill (1997, p. 546). Germany Code of Civil Procedure (Zivilprozessordnung), 10th Book, Bundesgesetzblatt I (1997) 3324, <http://www.uni-muenster.de/Jura.iwr/deutsch/berger/index.html> (English version). Code of Civil Procedure (Zivilprozessordnung), 8th Book, 5th Chapter (Interim Measures of Protection), Bundesgesetzblatt (1950) 533, <http://sunsite.informatik.rwth-aachen.de/Knowledge/germlaws/zpo/s916.html> (German version). Constitution of the Federal Republic of Germany (Grundgesetz), Bundesgesetzblatt (1949) 1, <http://www.uni-wuerzburg.de/law/gm00000_.html> (English version) Hong Kong Hong Kong Arbitration Ordinance as amended by the Hong Kong Arbitration (Amendment) Ordinance 1996, Chapter 341 of the Laws of Hong Kong, in Morgan (1997, p. 146). Hong Kong Arbitration (Amendment) Ordinance 1996 (75 of 1996), in Morgan (1997, p. 129). Basic Law of the Hong Kong Special Administrative Region of the People=s Republic of China, Decree No. 26 of the President of the People=s Republic of China (4 April 1990). Other countries Dutch Code of Civil Procedure (Wetboek van burgerlijke rechtsvordering), 4th Book (Arbitration), in A.J. van den Berg and P. Sanders (eds.), ICCA International Handbook on Commercial Arbitration, Vol. III, Supplement 7 (April 1987). Swiss Private International Law Statute (Gesetz ueber das Internationale Privatrecht), Chapter 12 (International Arbitration), in A.J. van den Berg and P. Sanders (eds.), ICCA International Handbook on Commercial Arbitration, Vol. III, Supplement 9 (September 1988). Italian Code of Civil Procedure (Codice di procedura civile), 4th Book, Title VIII (Arbitration), in A.J. van den Berg and P. Sanders (eds.), ICCA International Handbook on Commercial Arbitration, Vol. II, Supplement 17 (January 1994). Cases England Acrow (Automation) Ltd. v. Rex Chainbelt Inc. [1971] 3 All ER 1175. Anton Piller KG v. Manufacturing Processes Ltd. [1976] Ch. 55 (C.A.). Channel Tunnel Group Ltd. v. Balfour Beatty Construction Ltd. [1993] AC 334. The Mareva B Mareva Compania Naviera SA v. International Bulk Carriers SA [1980] 1 All ER 213. The Kostas Melas B SL Sethia Liners Ltd. v. Naviagro Maritime Corporation [1981] 1 Lloyd=s Rep 18. Union of India v. McDonnell Douglas Inc. [1993] 3 Lloyd=s Rep 48 and Yearbook Comm. Arb=n. XIX (1994) 235. Yousif v. Salama [1980] 1 WLR 1540 at 1542. Germany Bundesverfasssungsgericht, BVerfGE 57, 346 at 359. OLG Karlsruhe B Oberlandesgericht Karlsruhe, Zeitschrift fuer Zivilprozess International (ZZPInt.) 1 (1996) 91 at 93-94. LG Frankfurt B Landgericht Frankfurt, Neue Juristische Wochenschrift (NJW) 1983, 761 at 762-763. United States of America McCreary Tire & Rubber Co. v. CEAT SpA, 501 F.2d 1032 (3rd Cir. 1974) at 1038. Carolina Power & Light Company v. Uranex, 451 F.Supp. 1044 (N.D. California 1977) at 1052.