íkovo-Nagymaros Dam.61 Even if sustainable development is not in the nature of a legal obligation, it does represent a policy goal or principle that can influence the outcome of litigation and the practice of states and international organisations, and it may lead to significant changes and developments in the existing law.62 In that important sense, international law appears to require states and international bodies to take account of the objective of sustainable development, and to establish appropriate processes for doing so.
What these examples show is that subtle changes in the existing law and in existing treaties may come about through the application of such general principles. In Chapter 6 we will see how courts have made use of soft law principles in developing international law. In any system of law the ability to make changes on a systemic basis is important. How else could this be done in international law? New rules of customary law are not necessarily appropriate to the elaboration of such general principles and could not be created quickly enough; moreover, to take an obvious example, a treaty endorsing the precautionary principle would only bind the parties. A binding resolution of the UNSC may be a possible option, but only where questions of international peace and security are at stake. Thus, the consensus endorsement by states of a general principle enshrined in a soft law declaration is an entirely sensible solution to such law-making challenges.
Use of soft law techniques does not diminish the need for consensus among states, however. On the contrary, while amendments to treaties or implementing agreements may limp into force with only partial participation, adopting soft law principles without consensus support has little if any impact on the law-making process. The main advantage of adopting principles in soft law form is simply that the need to go through the process in treaty form is avoided. Once adopted, no ratification is necessary. This does not mean that soft law principles are an alternative to treaty amendment or implementation agreements in every case: far from it. Much will depend on what is proposed. But if the issue can be formulated in terms of interpretation or application of the existing terms of a treaty or other rule of law—as it could be in the case of the precautionary principle or sustainable development—there is no necessity to go further than soft law.
Resolutions of international organisations and multilateral declarations by states may also have effects on customary international law.63 Whether they provide evidence of existing law, or of the opinio juris necessary for new law, or of the practice of states, will depend on various factors which must be assessed in each case. A law-making resolution or declaration need not necessarily proclaim rights or principles as law, but as with treaties, the wording must be ‘of a fundamentally norm-creating character such as could be regarded as forming the basis of a general rule of law’.64 It is also obvious that declarations or resolutions setting out agreed norms or general principles ‘will usually have greater influence than recommendations’.65 The context within which soft law instruments are negotiated and the accompanying statements of delegations will also be relevant if assessing the opinio juris of states. Lastly, the degree of support is significant. A resolution adopted by consensus or by unanimous vote will necessarily carry more weight than one supported only by a two-thirds majority of states. Resolutions opposed by even a small number of states may have little effect if those states are among the ones most immediately affected.66 The attempt by the General Assembly in the 1970s to change the law on expropriation of foreign investments is a well-known example of the inability of majorities of states to legislate for minorities in this fashion.67 The General Assembly’s ban on deep seabed mining outside the framework of UNCLOS is another. In this case, the minority of objecting states maintained their own parallel regime, until eventually an agreement was reached.68
In an international system where the consent or acquiescence of states is still an essential precondition for the development of new law or changes to existing law, these examples show that opposing votes matter. Even if such resolutions can change the law for states which vote in favour, it is clear that they do not do so for the dissenting minority.69 Moreover, even consensus adoption will not be as significant as it may at first appear if accompanied by statements which seriously qualify what has been agreed, or if it simply papers over an agreement to disagree without pressing matters to a vote. For all these reasons, the adoption of resolutions by international organisations or of declarations by states should not be confused with law-making per se.
Some scholars have argued that General Assembly resolutions and declarations could create ‘instant’ customary law.70 Cheng concludes that the clearly articulated expression of opinio juris through the medium of a non-binding resolution or declaration may be enough, without further state practice, to afford evidence of a new rule of customary or general international law. For anyone seeking to use the UN General Assembly as a law-making instrument this is an attractive but generally unsustainable argument. Cheng himself rightly cautions against the facile assumption that UNGA resolutions make law, and his view of instant law-making is limited to very specific circumstances. First, it depends on a strong consensus in favour of such a resolution.71 Second, it requires appropriate wording. The principal UNGA resolutions on outer space fail this test, because they merely articulate principles by which states ‘should be guided’,72 rather than normative rules. Cheng was also writing before any of the leading modern ICJ cases on the creation of customary law were decided, and even his cautious formulation may now be too generous. The jurisprudence is not favourable, it must be said, to notions of instant law-making, but stresses instead the need for confirmatory practice, or at least the absence of contrary practice.73Moreover, although certain more appropriately worded UNGA declarations (for example the Universal Declaration of Human Rights and the Declaration on the Granting of Independence to Colonial Territories and Peoples) have undeniably had law-making effect, neither of these resolutions was adopted by consensus and it is doubtful that their impact was ever instantaneous.
If the resolutions on outer space did not make instant law, and were meant to be replaced by treaties,74 why adopt them at all? Apart from practical considerations of simplicity and speed of adoption compared to treaties, the importance of resort to the UNGA lies in the collective affirmation thereby provided for general rules of space law otherwise only impliedly asserted by the space states. The resolutions provided both a record of what all states believed the relevant rules should be, and evidence of opinio juris demonstrating the law-making significance of their earlier practice. As Brownlie observes, ‘In the face of a relatively novel situation the General Assembly provides an efficient index to the quickly growing practice of States.’75 Elsewhere he refers to the ‘decisive catalytic effect’ which such resolutions may have on state practice.76
In those circumstances it would be safe for space states to proceed on the assumption that there would be no opposition to activities conducted in conformity with the principles endorsed by the resolutions. That these principles were subsequently reaffirmed in treaty form shows both the value of soft law precedents as a prelude to later agreement on a more detailed international regime, and the preference for treaties as a means of stabilising the law within an appropriate institutional framework once the views and practice of states are settled. This may suggest a perception that soft law is too fragile an instrument to sustain the long-term regulation of common areas such as space or the deep seabed, but it is certainly an effective starting point when states need reassurance before commencing novel and previously unregulated activities.
The adoption of non-binding resolutions or declarations can also lead to changes in the existing law, in some cases quite quickly. The termination of drift-net fishing on the high seas is a good example of the successful use of UNGA resolutions in this way.77 Although the resolutions themselves have no legal force, and do not make ‘instant’ law, the widespread opposition to such fishing has been effective in pressuring many of the states involved to comply with the resolutions and phase out the use of driftnets. Today it seems unlikely that any state would wish to assert the right to use driftnets, although the practice has not yet been universally banned. Such changes in the law can of course only come about as a result of changes in practice by those states most closely involved. Moreover, even states initially voting against such resolutions may eventually conform to the general will. The initial opposition of colonial powers to UN resolutions on self-determination soon faded, and as the Western Sahara and East Timor cases show,78 former colonial states became the principal advocates of self-determination for their former colonies.
Why use non-binding resolutions for such purposes? Negotiating a global treaty on driftnets would have taken as long or longer; it would not have entered into force until there were enough ratifications, and if the relevant states failed to become parties they would not be bound anyway. In the latter case the law would not change unless these states changed their practice, which is no less true of the non-binding resolutions adopted by UNGA. In such cases, if the consensus for a change in practice is strong enough, a treaty is not necessary. If it is not strong enough a treaty will not necessarily strengthen it.
Banning a specific form of fishing is a relatively simple change in the law. Soft law instruments may not be as useful when more complex changes are needed. When it became apparent that the provisions of the 1982 UNCLOS relating to the high seas would not be sufficient to ensure sustainable fishing, an additional treaty was negotiated. Between the parties the 1995 UN Fish Stocks Agreement in effect elaborates and amends the 1982 UNCLOS. It is an important and far-reaching instrument that makes notable changes in the law on high seas fishing. Inter alia, freedom of fishing is confined to states which operate within the rules of regional fisheries organisations; fisheries law is given a newly environmental focus, emphasising sustainability and conservation of biological diversity and ecosystems; the enforcement powers of other states are extended, and regional agreements are brought within the UNCLOS dispute settlement system. However strong the consensus in favour of these changes, given their far-reaching nature and impact on other treaties, a treaty was clearly the only possible instrument in this situation. In reality, some of these measures have been controversial, and many fishing states remain non-parties to the 1995 Agreement, but resort to soft law would not have been any more effective in such circumstances. But if the weakness of soft law is that states are not obliged to comply, the same is no less true of an unratified or poorly ratified treaty. Whether the 1995 Agreement changes the law will still depend on how far it influences the practice of states, rather than on its binding force.
Soft law is manifestly a multi-faceted concept, whose relationship to treaties, custom and general principles is both subtle and diverse. At its simplest soft law facilitates progressive evolution of international law. It presents alternatives to law-making by treaty in certain circumstances, at other times it complements treaties, while also providing different ways of understanding the legal effect of different kinds of treaty. Those who maintain that soft law is simply not law have perhaps missed some of the points made here; moreover those who see a treaty as necessarily having greater legal effect than soft law have perhaps not looked hard enough at the ‘infinite variety’ of treaties, to quote Baxter once more. Soft law in its various forms can of course be abused, but so can most legal forms, and it has generally been more helpful to the process of international law-making than it has been objectionable.
We saw in Chapter 3 how the UNSC has begun to use its considerable power, and noted how some resolutions have a distinctly legislative character and purpose, despite obvious concerns about the Council’s legitimacy in this role. While General Assembly resolutions are at best only soft law, the legal effect of UNSC resolutions is more complicated. The Security Council has power to adopt decisions binding on all UN member states by virtue of Article 25 of the UN Charter. Not all provisions of UNSC resolutions are binding, however, including those which merely ‘recommend’, ‘call upon’ or ‘urge’ states to do or refrain from doing something. The wording, context and intent are all relevant, as is the UN Charter provision relied upon.79 ‘Decisions’ taken under Chapter VII of the Charter to maintain or restore international peace and security are the most important category of binding resolutions. A binding resolution may have legal significance in various ways, as we shall see below.
Are there any limits on the power of the UNSC to take binding decisions? Herein lies a profound debate on the constitutional limits of an admittedly powerful institution. It is not necessary to explore this debate in depth here,80 but some understanding of the issues is necessary because they impact directly on the law-making potential of the Council. The Council does not have absolute power; two obvious limitations are suggested by the wording of the Charter itself. First, Article 24 provides expressly that in carrying out its duties the Security Council ‘shall act in accordance with the Purposes and Principles of the United Nations’,81 while Article 25 refers to the obligation of member states to ‘accept and carry out the decisions of the Security Council in accordance with the present Charter’. Delbruck gives this phrase a narrow reading limited to procedural compliance with the Charter,82 but as Martenczuk argues, the ICJ’s treatment of the Lockerbie cases suggests that it has ‘resisted all attempts to remove Chapter VII of the Charter from the ambit of legal interpretation’.83 It might thus be argued that the Council’s assumption of a more general law-making authority under Chapter VII falls outside the substantive scope of its powers as intended by the negotiators of the Charter.84More probably, however, the purposive reading of the Charter adopted by the ICJ in the Certain Expenses case85 would lend support to the Council’s law-making activities so long as they can reasonably be related to the maintenance or restoration of international peace and security under Article 39.86 In Chapter 3 we saw how the Council has made generous use of Chapter VII, asserting the right to act even in respect of internal humanitarian crises, civil war, governmental breakdown or racist oppression. In some of these cases the threat to ‘international peace and security’ is tenuous. Given the breadth of the discretion afforded to the Council, however, it is not surprising that judicial decisions have shown considerable deference to its political judgments and treated its decisions as presumptively valid.87Nevertheless, it is possible that the legality of a resolution could be challenged on the ground that it is far removed from peace and security.
A second limitation may flow from the concept of jus cogens. Some of the Council’s resolutions relating to the war in Bosnia have been challenged on this basis because, it was argued, they facilitated genocide by denying Bosnia the right to defend itself.88 This argument would at least support the conclusion that Council resolutions cannot legitimise torture, genocide, war crimes or any other activity falling within the narrow category of jus cogens presently supported by judicial decisions. In the context of possible Security Council action against international terrorism these may prove to be significant limitations.
More questionably, it might be argued that in using its powers the Council must respect general international law. A possible example is the Tadic case, considered in Chapter 3, which addresses human rights limitations on the Council’s power to establish criminal tribunals. However, interpreting and applying UNSC resolutions in accordance with human rights law, or with other relevant rules of international law, does not tell us that the Council cannot lawfully pass resolutions inconsistent with general international law; it merely compels it to do so in appropriate terms. More pertinently, it has been suggested that the Council cannot force states to agree to third party settlement of disputes, or impose territorial boundaries, or compel states to extradite suspected terrorists, because to do so would be inconsistent with existing law, procedural due process or treaty commitments.89 If that were correct, then the Council’s power to legislate generally about such matters would be similarly constrained. It is doubtful that this is a good argument, however. In UNSC 687 the Council carefully refrained from imposing a boundary on Iraq, but it is not clear that its reasons for doing so were constitutional rather than political.90 In Lockerbie incompatibility with an existing treaty did not persuade the ICJ to restrain implementation of a UNSC resolution regarded by the court as prima-facie valid.91 Moreover, if the Council is bound to act within existing international law then its power to act at all will be severely compromised and open to challenge in most, if not all, cases. Why, for example, should a state which claims to be acting lawfully in self-defence pay any attention to a resolution imposing a ceasefire? Why should it evacuate territory it believes to be lawfully its own, however implausibly? Why should it do anything that general international law does not already require or empower it to do? On that basis the Council would find itself constantly embroiled in endless argument about the legality of its decisions, with serious consequences for its effectiveness. This outcome is a recipe for emasculation, and as we shall see in Section 3.2 below it is inconsistent with Article 103 of the UN Charter as interpreted by the ICJ, and unlikely to appeal to any international tribunal. In his classic work on the United Nations, Kelsen concluded that ‘The Charter does not provide that decisions . . . in order to be enforceable must be in conformity with the law which exists at the time they are adopted.’92
While the precise limits on the Council’s power of decision remain uncertain, and the procedures for challenging the constitutionality of its decisions are unsatisfactory,93 it is nevertheless clear that at a minimum there must be some threat to international peace and security before the Council can take binding decisions under Chapter VII. This might justify action to restore the environment, or ensure democratic elections,94 but it is less likely that it would cover measures designed to save the whale, or to secure education rights for indigenous peoples, for example. The next section shows that the Council’s resolutions have greater legal effects than the General Assembly’s but it is already clear that it can only exercise its powers within narrower limits.
Valid and binding decisions of the UNSC not only affect all states, the jurisprudence suggests that they also over-ride inconsistent international law. Customary law will necessarily yield to binding UNSC decisions, if that is the intention and if the resolution is so worded, on the straightforward basis that parties to a treaty are free to contract out of custom and will in this case have done so under Article 25 of the Charter. Furthermore, under Article 103 of the Charter a valid and binding UNSC decision will also prevail over inconsistent international agreements. This was the clear view of the ICJ in the Lockerbie case.95 The Court refused an application by Libya for provisional measures on the ground that, prima facie, resolution 748 requiring Libya to surrender terrorist suspects for trial was binding and thus prevailed over the inconsistent provisions of the 1972 Montreal Convention for the Suppression of Unlawful Acts against the Safety of Civil Aviation. The Court left the matter open for reconsideration on the merits only in respect of the question whether resolution 748 was indeed binding. Moreover, reviewing UNSC resolutions which expressly or by implication assert priority over other treaties, Bernhardt concludes that ‘the principle that binding SC decisions taken under Chapter VII supersede other treaty commitments seems to be generally recognised . . . ’.96
The significance of this conclusion cannot be over-estimated. Unlike any other international organisation, it gives the UNSC the power to rewrite or dispense with existing international law in particular situations, and possibly in more general terms.97 Potentially, Security Council resolutions may thus have as great or greater significance than the concept of jus cogens. In effect the capacity to override other treaties and general international law amounts to a claim to formal legislative capacity. While challenges to their validity may be a tenable response, it is neither possible for states to opt out of binding UNSC resolutions, nor to formulate reservations. Nor, since they are not treaties, is it clear on what basis UNSC resolutions should be interpreted.98Used vigorously, and on the basis of the liberal interpretation it has given to ‘peace and security’, the Council could make immediate, mandatory, and enforceable changes to international law. In this respect its resolutions are very different from the soft law adopted by the General Assembly. This is not in itself an objectionable outcome and it is one which promotes coherence in international law-making; what matters more, as we saw in Chapters 1 and 3, is the legitimacy of the process, not the capacity to over-ride existing law.
Treaties are necessarily the product of negotiation, and their legal force as treaties depends entirely on the consent of the parties, variously expressed by signature, ratification, accession or in any other agreed way. A party to a treaty is bound to comply with its terms and entitled to act in accordance with them; a non-party is not, unless the parties to the treaty intend to create rights or obligations for third states.99 Viewed simply from the perspective of the law of treaties, therefore, treaties are a source of rights and obligations for the parties; most were never intended to be law for all states. Whether they are contracts, conveyances or constitutions, the vast majority are thus of no further significance for a work on international law-making.100
Paradoxically, however, much of the most important international law-making in the modern world is undertaken in treaty form. There are several good reasons for this development. As we saw in Chapters 3 and 4, both the process of negotiation and the binding character of treaties have made them the closest analogy to an international legislative instrument so far devised, even if the analogy is inexact. The potential for treaties to become sources of international law was acknowledged as early as 1920 in Article 38 (1) (a) of the Statute of the Permanent Court of International Justice, insofar as they establish ‘rules expressly recognised by the contesting States’. The UN Charter, the 1949 Geneva Conventions on humanitarian law, the 1969 Vienna Convention on the Law of Treaties, the 1982 UN Convention on the Law of the Sea, and the 1994 WTO Agreement all illustrate such rules. Thus the idea that some treaties are ‘law-making’, or ‘traités-lois’, is neither new nor controversial, but it poses some distinct and difficult questions that we try to address in the following sections.
It needs to be appreciated, moreover, that a treaty may be ‘law-making’ in several different senses. At its simplest, a treaty such as the UN Charter is effectively law for all states because it sets out important general rules, powers and principles and nearly every state is a party. Such universal participation is unusual, and the Charter is a rare example. However, as we saw in Chapter 3, there are many other multilateral treaties which, even if they do not enjoy universal participation, nevertheless create regimes which are in force and effectively law for a large number of states parties. These regimes need to evolve in response to new policies and problems, and as we shall see below there are various mechanisms for achieving this.
Even where participation is much less than universal, a multilateral treaty may nevertheless affect other states because certain provisions either confer rights on ‘all states’ or through practice or intent have become accepted by parties and non-parties alike as the basis of new customary law. On the other hand there are also a significant number of multilateral treaties with few parties, sometimes many years after adoption. These agreements remain ‘limping’ for various reasons:101 some simply do not meet real needs; others may encounter domestic opposition; a few are merely symbolic. This does not mean that poorly supported treaties cannot influence the emergence of customary law—a point excellently illustrated by the Vienna Convention on Treaties itself—but the reasons for poor ratification may also limit their wider impact. Finally, a treaty may afford evidence of existing law, as we saw in our discussion of codification in Chapter 4.
A treaty does not ‘make’ customary law, but like soft law it may both codify existing law and contribute to the process by which new customary law is created and develops. The process has been fully explored by the ICJ in two cases, the North Sea Continental Shelf case and the Nicaragua case. The approach taken by the court is subtly different in both cases. As we saw in Chapter 4, in the North Sea case the Court accepted that a normatively worded provision of a multilateral treaty could contribute to the formation of a new rule of customary law if the subsequent practice of a sufficiently widespread and representative selection of non-parties conformed to the treaty and there was additionally evidence of opinio juris.102 In the Nicaragua case, the ICJ reiterated that ‘the shared view of the parties as to the content of what they regard as the rule is not enough. The Court must satisfy itself that the existence of the rule in the opinio juris of states is confirmed by practice.’103However, that practice need not be perfectly consistent nor conform rigorously in order to establish its customary status, provided inconsistent conduct is treated by the states concerned as a breach of the rule, not as an indication of a new rule. Attempts to justify inconsistent conduct serve, on this view, merely to confirm the rule in question. In this context only conduct amounting to an outright rejection of the alleged rule will constitute genuinely inconsistent practice. The Nicaragua case also recognised that the embodiment of a rule in a treaty provision (in this case the UN Charter) does not displace an existing rule of customary international law or prevent its continued development.104
The Court’s findings of law in Nicaragua do not rest only on the normative impact of the UN Charter. The point has already been made that the opinio juris of states was also evidenced by consensus adoption of comparable resolutions in the UN General Assembly. Although the Court is careful to avoid any suggestion that custom can be established simply by states declaring the law in treaties and soft law resolutions, it comes close to doing so. The only significant caveat is that there must be no inconsistent state practice. This reverses the approach taken in the North Sea case, where the Court emphasised the need for consistent state practice. The circumstances of the two cases are different, however. In North Sea, the supposed equidistance rule was not acceptable to a number of states, and certainly not to Germany, nor was the practice of non-parties to the treaty consistent. By contrast, both parties in the Nicaragua case consistently expressed support for the same rule on the use of force, even in the face of flagrant violations. Moreover, it should not be forgotten that the United States is a party to the UN Charter, even if in the circumstances of the Nicaragua case the Court was precluded from applying the Charter to the dispute. Finally, the use of force in violation of international law contravenes a norm of jus cogens. By definition, such rules are recognised by the international community as whole. The delimitation of seabed boundaries on the basis of equidistance is not in the same category.
In the Malta/Libya Continental Shelf case105 the International Court had the opportunity to consider once again the effect of a law-making treaty on the evolution of customary law relating to seabed boundaries. The 1982 UNCLOS was not at that time in force, nor were Malta or Libya parties, but both had participated in its negotiation and were signatories. Both states accepted for the purposes of the litigation that ‘some of its provisions constitute, to a certain extent, the expression of customary international law in the matter’.106 The Convention had introduced the new concept of the 200-mile exclusive economic zone (EEZ), conferring on coastal states sovereign rights over seabed minerals and living resources within that area, independently of the geographical continental shelf. The question before the Court was: what legal principles now applied to delimitation of the parties’ shared continental shelf in customary law? The Court saw no difficulty in holding that the Convention had changed the law. Although neither Libya nor Malta claimed an EEZ, the practice of other states based on the Convention demonstrated that the EEZ had become part of customary law;107 the Convention showed how the EEZ and continental shelf are inter-linked in modern international law, and the development in the customary law of the continental shelf is reflected in Articles 76 and 83.108
Here we can see how novel provisions of a convention negotiated using the consensus/package deal procedure discussed in Chapter 3 and adopted by ‘an overwhelming majority of states’109 are rapidly translated into law on the basis of state practice—and something more. Because of course there was no state practice on the relationship between the 200-mile EEZ and the delimitation of the continental shelf: that relationship could only be inferred from the provisions of the Convention itself. In effect, once state practice based on the Convention had endorsed the EEZ regime, the Court had to accept that consequential changes in the customary law of the continental shelf law resulted. The only basis for doing so was the Convention itself. Hence the attitude of both states to the treaty, and the general acceptability of the negotiated text, were important elements, as in the Nicaragua case.
What conclusions about the normative force of law-making treaties can we draw from these three cases? First, the Court had no difficulty accepting that treaties ‘may have an important role in recording and defining rules deriving from custom, or indeed in developing them’.110 Second, the cases accept that there is a law-making intention behind the negotiation of certain multilateral treaties. This can constitute evidence of opinio juris in favour of new general rules of international law, especially if the treaty was negotiated by consensus or has the consistent support of a large majority of states. Third, we can see that support for a treaty rule, however universal, cannot by itself create ‘instant’ law. Such treaties will only create new law if supported by consistent and representative state practice over a period of time. That practice can in appropriate cases consist mainly of acquiescence, or the absence of inconsistent practice. How long a time is required will depend on the circumstances. It may be very short indeed if the subsequent practice is widespread and consistent, as in the case of the exclusive economic zone, or if the treaty is a codification treaty, as we saw when considering the Gabíkovo case.111 Fourth, it evidently matters little whether the treaty is in force, or widely ratified. What is most important is simply that the more widely supported the treaty text is shown to be, the easier it will be to establish its law-making effect. The longer it takes to establish consistent opinio juris or consistent state practice the harder it will be to establish a new rule of customary law.
This leads on to a final question: is the Court right in these cases to place so much emphasis on state practice? Could it be that the adoption of a consensus text is itself enough, or should be? Let us take another example from the 1982 UNCLOS. Article 218 gives port states express power to investigate and prosecute discharge violations wherever they have taken place. This power covers both high seas offences, and violations within the coastal zones of another state, although in the latter case the port state may only act in response to a request from the state concerned. In effect Article 218 has created a form of universal jurisdiction based on the presence of the vessel within the port.
It is, however, a novel development in the law of the sea to confer jurisdiction on port states in this way. Although the Lotus