Codification and Progressive Development of International Law
It has been suggested that some of the articles have ‘a bootstrapping quality, helping to shape the law to match the draft’,93 but this tells us only that any attempt to set out a systematic account of the law built largely on international judicial decisions and awards by claims commissions cannot be a mere codification of existing rules. In significant respects it was far from clear what those rules were, still less what they should be. The prolonged debate on countermeasures, erga omnes obligations, the role of fault, and the definition of ‘injured state’ are only some of the elements where the Commission’s work can be said to develop the law rather than merely codify it. Indeed some went so far as to characterise the Commission’s work as an attempt to construct a ‘system of multilateral public order’ rather than codifying the law of state responsibility.94 Others criticised it as mostly a backward-looking endeavour which failed to address the expanding scope of contemporary international law or the emergence of non-state entities as significant actors. Some saw the whole endeavour as flawed because it presumed that a unified body of law could cater for a multiplicity of special regimes. Crawford argues that the articles are neither as inflexible nor as conservative as some of these comments might suggest, and that the level of controversy is ‘overstated’.95 Nor is the process finished: ‘The articles will have to prove themselves in practice, and that is a process which will require careful assessment. Indeed, the point of the ILC’s recommendation, in the first place at least, was to allow such a process of testing and assessment to continue . . . on a case-by-case basis.’96 Here we can see the Commission itself pointing to the very important impact which ICJ decisions and those of other international courts have on the status of ILC draft articles and conventions as customary law.97
It is not necessary here to enter further into an account of the ILC draft articles on state responsibility.98 Whatever their merits may be, they show the value of an expert body which can give systematic shape to legal principles of quite fundamental importance to the character and scope of contemporary international law. In this case, unlike some others that we examine below, it cannot be said that the Commission failed to rise to the challenge, even if some might have preferred it to go further. It is not surprising that the draft articles on state responsibility have been referred to as one of the ILC’s most significant accomplishments.99
In 2004 the UN General Assembly adopted a Convention on the Jurisdictional Immunities of States and their Property.100 This was not the first attempt to codify international law relating to state immunity. Draft articles were adopted by the Institut de Droit International in 1891 and in 1927 the Committee of Experts on Progressive Codification of International Law recommended to the League of Nations that the subject was ripe for codification, in response to which a convention of 28 articles was drafted by American lawyers as part of the Harvard research programme.101 None of these projects came to anything, but in 1949 the subject reappeared on the list of topics provisionally selected by the ILC for codification. Not until 1978 did the Commission decide to proceed with the topic, and the draft articles it adopted in 1991 proved controversial and failed to reconcile the significant differences between states. It took a further 13 years of negotiations at the UN to produce a partially revised text that was finally adopted without a vote.
Whether the new convention will become accepted as a codification of international law on the subject cannot at this stage be certain. Its impact will depend significantly on how far it succeeds in encouraging states to harmonise existing national laws. It is unlikely to change the national law of non-parties and to that extent it will be important that as many states as possible become parties: mere acquiescence alone will not suffice. Moreover, the Convention is not a package deal and reservations are not prohibited. So long as they do not undermine the object and purpose of the Convention, it will be open to individual states to pick and choose those elements which they are willing to accept while reserving others. There is thus no guarantee that national laws will be harmonised even among the parties unless reservations are kept to a minimum. Even then, there remain uncertainties in a text which may be interpreted in significantly different ways.102 Finally, the important question of immunities with respect to proceedings brought by victims of torture, crimes against humanity, or other breaches of human rights obligations remains unresolved, nor is immunity from criminal prosecution covered by the Convention. In all of these respects the customary international law will continue to evolve, if at all, on the basis of national law and judicial decisions, although the difficulty of resolving conflicting trends through adjudication is apparent in the case law, national and international, that has developed during the past decade.103
Nevertheless, even if it is not comprehensive and may not fully harmonise national laws, the adoption of a global convention on state immunity is a significant achievement. It articulates for the first time a common international standard of limited immunity based largely on the practice of those jurisdictions, such as the United Kingdom, whose most recent legislation has sought to ensure that foreign states can be sued in respect of commercial contracts and other essentially non-sovereign transactions. There is now agreement on the exceptions to immunity, both from suit and from execution of judgments or actions in respect of state property, and on what constitutes the ‘state’ for this purpose. For a number of civil law countries it will provide a valuable means of modernising national law and the immunity provisions of their civil codes. For states such as China which had formerly adhered on ideological grounds to the doctrine of absolute immunity from suit, the Convention offers the opportunity to move to a common conception of immunity limited to the exercise of sovereign authority.104
The 2004 Convention thus addresses the main problems which had originally persuaded the ILC to start work on the topic and on which the majority of states had from the beginning sought agreement. Faced with widely differing national laws on the subject, and no clear consensus on what immunities international law either permitted or required, the working group which recommended the choice of topic in 1978 concluded that ‘it was in the interest of states generally that the rules of international law governing state immunities should be made generally more ascertainable so as to give general guidance to states for the adoption and maintenance of a consistent attitude . . . ’.105 However, the ILC clearly underestimated the extent to which strong national interests would impede the special rapporteur’s attempts to limit the immunity of states, and the ideological differences with regard to the personality, capacity and functions of the state. The result was a draft convention adopted by the Commission in 1991, which was unsatisfactory to many states in various ways.106 It delegated the distinction between commercial and non-commercial transactions largely to national law. It also failed to reconcile broader and narrower conceptions of what constituted the state or how far states could be held liable for the commercial transactions of state enterprises, and it left states with extensive immunity from any form of execution of judgments. Moreover, the end of the cold war found former communist states with extensive state-owned enterprises uncertain how to proceed with a draft that could expose them to liabilities from which they might no longer be immune. The impact of privatisation and economic globalisation made harmonisation of the law on state immunity more important, but 1991 was not the right moment politically to proceed to do so. Not surprisingly, when the ILC sent its draft articles to the 6th Committee for consideration, they proved too politically controversial. Nor could the 6th Committee itself resolve differences of view among states on the most fundamental issues; consultations undertaken in 1993/4 failed to agree on solutions. Indeed participants could not even agree on whether the differences could be resolved.
The ILC cannot be blamed for failing to foresee the end of the cold war and the collapse of communism, or for the prolonged inability of the 6th Committee to reach agreement. As the Italian government observed, ‘the ILC draft was the result of a study which had been carried out during a period in which the institutional and economic landscape were far different from now’.107 Nevertheless, apart from its understandable inability to reconcile the irreconcilable, the Commission’s difficulties with state immunity betray other failings of a more systemic kind. Hafner, who chaired the 6th Committee working group and the ad hoc committee which eventually negotiated the final text of the Convention, has pointed to the absence of any interaction or exchange between the ILC and the 6th Committee.108 He confirms the point made earlier that lack of response from states left the ILC inadequately informed about what would be acceptable,109 and concludes that the topic shows the lack of a proper relationship between states and the ILC. Politics alone do not explain all the problems with the ILC’s work on state immunity; deficiencies of process also play a role.
Ultimately, the problems left unsolved in 1994 were finally dealt with ten years later, after further negotiations in a 6th Committee working group and an ad hoc committee.110 These bodies took into account further comments from states, including China, Japan and Italy, and also had the benefit of proposals for revision from a working group of the ILC, which the 6th Committee had requested. Significant revisions to the ILC draft resulted, most notably in the definitions of ‘State’ and ‘commercial transaction’, in the article on contracts of employment, and in articles limiting the immunity of state property from pre- and post-judgment measures.111 The ad hoc committee also agreed on a dispute settlement clause and an annex setting out interpretations of certain articles. In effect this committee became a negotiating conference of states. Comments made by China in 2001 in support of the revised ILC draft show the extent to which political change made legal change easier.112 Thus, when finally adopted, the UNGA resolution was able to note ‘the broad support for the conclusion of a convention on jurisdictional immunities of States and their property’.
Probably the most important conclusion to draw from the codification of international law on state immunity is that timing is crucial. For over a century successive bodies had concluded that the time for codification was ripe; the experience of the ILC appears to show that for most of this period it was never ripe, and with hindsight it seems unsurprising that nothing was achieved for so long. If the ILC was wrong in its timing, it nevertheless produced a draft that offered a substantial basis for further negotiation and eventual agreement between states when the time did become very ripe only a few years later. Most of its original text remains in the new Convention unchanged or changed only to a limited extent. If the 1991 version was unsatisfactory in some respects, the Commission’s work had succeeded in highlighting the central areas of disagreement.
Viewed as part of a larger process of international law-making and harmonisation of very disparate national laws, this is not an insignificant achievement. It is not realistic to expect treaties to emerge from the ILC fully formed, universally supported and wholly unblemished by ambiguities or inelegant compromises. As we emphasised earlier, the ILC is an expert body of international lawyers, not a drafting committee of states. Faced initially with strong ideological differences surrounding the topic, and then with serious changes in the political landscape, it is to be expected that only significant political involvement in the process could secure ultimate consensus. Of course, earlier input from governments would have helped the Commission determine what states really thought, but this presupposes that states had fully formed views on the topic that they could have shared with the Commission during its work rather than afterwards. This may be an optimistic view of governments, and it discounts the possibility that the ILC’s work helps the views of states to evolve, even if the evolution may be in contrary directions. The notion that some element of dialogue exists between the Commission and states is not entirely fanciful, even if it is imperfect and sporadic. Certainly, without such a dialogue codification projects cannot be expected to prosper, as we saw earlier. Lastly, it is also worth noticing the absence of other participants in a dialogue which has important implications, inter alia for remedying breaches of human rights law and international criminal law. No doubt the involvement of human rights NGOs would have made agreement on the convention text even harder to achieve, but their influence will nevertheless be felt at national level and that may now make ratification harder for some states.113
No other topic has occupied the ILC for longer; few ILC conventions have established such a universally accepted, comprehensive and enduring codification than the 1969 Vienna Convention on the Law of Treaties; none has been so widely relied on by governments and courts. By any measure, this is one of the ILC’s successes, despite a relatively poor record of ratification of or accession to the 1969 Convention and the even worse record of the 1986 Vienna Convention on the Law of Treaties between States and International Organisations.114 Moreover, even for those states which are parties, the Vienna Conventions do not apply to all treaties. Neither Convention is retrospective, so prior treaties continue to be governed by customary law, as do treaty relations between parties to the Vienna regime and non-parties. Thus the real significance of the ILC’s achievement lies not in the conventions themselves, but in the extent to which they have successfully become accepted by international and national courts, or by governments and foreign ministries, as restatements of customary law.
The 1969 Vienna Convention is not solely the work of the ILC. As a draft convention it was submitted to a UN diplomatic conference that undertook some significant revision and redrafting. The most notable addition is Article 66 providing for compulsory dispute settlement procedures in respect of Part V of the Convention, although ironically these procedures have never been used. Without this compromise text it was likely that the diplomatic conference would have failed.115 Articles 11 (consent to be bound), 13 (exchange of instruments), 46 (2) (violation of internal law), 60 (5) (material breach of humanitarian treaties) and 74 (diplomatic and consular relations) were also proposed by states rather than by the ILC. The 1969 Convention and its 1986 counterpart, with their subtle mélange of codification and progressive development, have undoubtedly shaped the modern law of treaties,116 but it is the interaction of the ILC, the diplomatic conferences, subsequent practice and judicial decisions which have brought this about.
While building on existing law, the 1969 Convention nevertheless made important clarifications, reformulations and additions, most notably in regard to interpretation, invalidity, reservations and jus cogens. Some provisions go beyond previous precedents, such as Article 18 on the obligation not to defeat the object and purpose of a treaty pending signature, or Article 20 on the effect of objections to reservations. The degree of progressive development contained in Part V of the Convention, dealing with invalidity, termination and suspension, was acceptable to many states only because Article 66 provides for compulsory settlement procedures in the event of a dispute.117 Nevertheless, the ICJ has been willing to accept many of these articles as evidence of current customary law, and thus applicable to states not bound by the Convention. Article 62 on change of circumstances was applied by the ICJ in the Icelandic Fisheries cases,118 although the Convention was not then in force and had been adopted only four years previously. Similarly, Article 60 on breach of treaty was relied on in the Namibia Advisory Opinion 119 and in the ICAO Council case.120 Most notable of all is the ICJ’s reliance on various articles from Part V of the 1969 Convention in the Gabcíkovo case,121 notwithstanding that the 1977 treaty at issue in that case pre-dated entry into force of the Vienna Convention. A few entirely new provisions, such as Article 53 (on jus cogens), were also controversial for some states. Notwithstanding these objections, however, the concept of jus cogens has acquired a significant pedigree in case law, although none of the judicial decisions deals with the specific issue of validity of a treaty that contravenes such a norm.122
Indeed, it can be said that on most questions both conventions have come to be widely, if not universally, accepted as authoritative statements of the customary law of treaties. On that basis the 1969 Convention’s provisions on interpretation (Articles 31 and 32) have been applied in the case law of nearly all international tribunals and many national courts.123 Commenting on the Gabíkovo case, Aust concludes that it is not unreasonable to assume that the International Court will apply the same approach to virtually all of the provisions of the 1969 Convention, and as he points out there is no case in which the Court has found that the Convention does not reflect customary law.124
Only in one respect has the 1969 Vienna Convention failed to generate consensus on the applicable law, but the failure is illuminating. With one important modification,125 the ILC’s proposals on reservations to treaties were adopted as Articles 19–23 of the 1969 Convention, but they have never become fully established. When the permissibility of reservations was considered in the Reservations to the Genocide Convention Advisory Opinion,126 the ICJ held only that states were free to make reservations to humanitarian conventions, provided that the reservation was compatible with the object and purpose of the convention, and subject to the right of other parties to object. This departure from the then prevailing practice which disallowed reservations except with the consent of other parties was justified in the interests of promoting wider participation in humanitarian treaties. The ILC adopted the Court’s approach and with some amplification applied it to all treaties in its draft articles on reservations, which subsequently became Articles 19–23 of the 1969 Vienna Convention. In the Anglo-French Continental Shelf Arbitration 127 the arbitrators generally followed the Vienna Convention provisions when determining the permissibility and effect of reservations in customary law.
However, in a number of later cases it became apparent that human rights tribunals and treaty-monitoring bodies were taking a different approach to these questions.128 Instead of leaving it to states in accordance with Vienna Convention rules to object to or acquiesce in reservations, and to treaty relations inter se, tribunals and monitoring bodies were deciding questions relating to the permissibility and effect of reservations themselves. Reservations determined to be impermissible were in some cases disregarded and in others no objections to permissible reservations were allowed. Moreover, as several studies observed, neither the Vienna Convention nor state practice had fully clarified the consequences of making an impermissible reservation, regardless of the category of treaty concerned.129
Given such divergent views, and continuing uncertainty over other matters, such as the distinction between reservations and interpretative declarations, the ILC once more returned to the topic of reservations in 1993. It envisaged drafting a guide to practice for states and international organisations rather than a new or amended treaty.130 The resulting guidelines are essentially an amplification of the Vienna regime, clarifying uncertainties or filling in gaps. In this form the Commission’s draft has been the subject of very extensive consultation with states and international organisations.131Nevertheless, one reason why the topic has proceeded so slowly is that views remain divided and the special rapporteur has been reluctant to allow matters to come to a premature conclusion.
This is particularly true with respect to the problem of impermissible reservations to human rights treaties, which the special rapporteur did not begin to address until 2005.132 In 1994 the UN Human Rights Committee had explicitly found that ‘the provisions of the Vienna Convention on the role of State objections are inappropriate to address the problem of reservations to human rights treaties’.133Despite this, the ILC provisionally concluded in 1996 that the Vienna Convention was intended to afford a single legal regime for reservations, and that this regime remained suited to all types of multilateral treaty, including human rights treaties.134