The distinction between bilateral (reciprocal) and multilateral (integral) obligations cannot, in so many words, be found in any codified rule of international law. Nonetheless, it has major consequences and is reflected in a number of norms on the law of treaties and state responsibility.
In the Reservations to the Genocide Convention case (1951), the ICJ gave particular importance to the "objects" of the Convention. It noted that "[t]he Convention was manifestly adopted for a purely humanitarian and civilizing purpose. It is indeed difficult to imagine a convention that might have this dual character to a greater degree, since its object on the one hand is to safeguard the very existence of certain human groups and on the other to confirm and endorse the most elementary principles of morality".1 With reference to these "objects", the ICJ lay the foundation of what was to become the distinction between reciprocal and integral obligations:
"In such a Convention [as the Genocide Convention] the contracting States do not have any interests of their own; they merely have, one and all, a common interest, namely, the accomplishment of those high purposes which are the raison d'être of the convention. Consequently, in a convention of this type one cannot speak of individual advantages to States, or of the maintenance of a perfect contractual balance between rights and duties. The high ideals which inspired the Convention provide, by virtue of the common will of the parties, the foundation and measure of all its provisions".2
It was with reference, inter alia, to these "objects" that the Court made its main finding: "The object and purpose of the Convention thus limit both the freedom of making reservations and that of objecting to them. It follows that it is the compatibility of a reservation with the object and purpose of the Convention that must furnish the criterion for the attitude of a State in making the reservation on accession as well as for the appraisal by a State in objecting to the reservation".3 The ICJ's approach to treaty reservations was incorporated subsequently in Art. 19(c) of the VCLT, prohibiting reservations to a treaty that are "incompatible with the object and purpose of the treaty".
In earlier opinions by individual PCIJ judges, other hints at a distinction between reciprocal and integral treaties were made. In the Customs Régime Between Germany and Austria case, Judge Anzilotti questioned whether the parties to the 1922 Geneva Protocol
"were in a position to modify inter se the provisions of Article 88 [of the Treaty of Saint-Germain], which provisions ... form an essential part of the peace settlement and were adopted not in the interests of any given State, but in the higher interest of the European political system and with a view to the maintenance of peace".4
Another reference can be found in the dissenting opinions of judges Van Eysinga and Schücking in the Oscar Chinn case. In contrast to the majority of the PCIJ, judges Van Eysinga and Schücking expressed the view that the 1919 Convention of St. Germain relating to the Congo Basin was void between its signatories on the ground that it modified the earlier General Act of Berlin of 1885 without the assent of all the signatories thereto. Judge Van Eysinga expressed it thus:
"the Berlin Act presents a case in which a large number of States, which were territorially or otherwise interested in a vast region, endowed it [the Congo Basin] with a highly internationalized statute, or rather a constitution established by treaty, by means of which the interests of peace, those of `all nations' as well as those of natives, appeared to be most satisfactory guaranteed ... [It] does not create a number of contractual relations between a number of States, relations which may be replaced as regards some of these States by other contractual relations ... This régime, which forms an indivisible whole, may be modified, but for this agreement of all contracting Powers is required".5
Fitzmaurice refined the distinction between treaties referred to in the previous section and re-phrased it as one between "reciprocal" or "concessionary" obligations, on the one hand, and "integral" obligations, on the other. Multilateral treaties of the "reciprocating type" are those "providing for a mutual interchange of benefits between the parties, with rights and obligations for each involving specific treatment at the hands of and towards each of the others individually".6 Whereas multilateral treaties of the "integral type" are those "where the force of the obligation is self-existent, absolute and inherent for each party".7 In other words, "integral obligations" are those "towards all the world rather than towards particular parties"8 and "do not lend themselves to differential application, but must be applied integrally".9
The standard example given by Fitzmaurice of a treaty of the reciprocating type was the 1961 Vienna Convention on Diplomatic Relations; that of the integral type, the 1948 Genocide Convention.
Fitzmaurice attached two important legal consequences to this distinction, one in the field of termination/suspension of treaties, the other in the field of conflict between treaties. Treaties of the reciprocating type could, in Fitzmaurice's view, be suspended or terminated as a result of fundamental breach.10 Moreover, later treaties conflicting with previous ones of the reciprocal type were, in his view, not null and void (instead, priority rules applied).11 Integral treaties, in contrast, could, under Fitzmaurice's draft, not be terminated or suspended by the other parties as a result of breach ("the juridical force of the obligation is inherent, and not dependent on a corresponding performance by the other parties to the treaty").12 In addition, any subsequent treaty concluded inter se by the parties to such integral treaty which "conflicts directly in a material particular with the earlier [integral] treaty will, to the extent of the conflict, be null and void".13
Fitzmaurice also added a third type of multilateral treaties, namely those of an "interdependent nature", where "the participation of all the parties is a condition of the obligatory force of the treaty".14 He gave treaties on disarmament as an example of interdependent treaties. In terms of termination/suspension as a result of breach, interdependent treaties could, in Fitzmaurice's view, be terminated in their entirety by the other parties in case of fundamental breach (not just suspended or terminated partly as was the case for reciprocal treaties), since for these treaties "performance by any party is necessarily dependent on an equal or corresponding performance by all the other parties".15 However, much like "integral treaties" (and unlike "reciprocal treaties"), a later inter se treaty which "conflicts directly in a material particular with the earlier [interdependent] treaty will, to the extent of the conflict, be null and void".16 This notion of "interdependent treaties" will not be further referred to here. Unlike the notions of reciprocal and integral treaties, the concept of interdependent treaties has not been generally used as a distinct category subsequently to Fitzmaurice's reports.17 For purposes of conflict of norms (essentially, the legality of inter se modifications), these interdependent treaties can, indeed, be equated with integral treaties.
Fitzmaurice's distinction between reciprocal, integral and interdependent treaties was not maintained in the Vienna Convention as it was finally concluded. Nonetheless, it left its traces in not less than six different provisions. The Convention deals with termination/suspension as a result of "material breach" in its Art. 60 and conflict with earlier treaties in its Arts. 30, 41, 53, 58 and 64.
Firstly, under Art. 60.5 termination/suspension as a result of material breach is not allowed in case of "provisions relating to the protection of the human person contained in treaties of a humanitarian character, in particular provisions prohibiting any form of reprisals against persons protected by such treaties". These treaties can, indeed, be seen as an example of integral treaties in respect of which Fitzmaurice precluded termination and suspension18 (although not all integral treaties have been kept outside the scope of Art. 60, contrary to what Fitzmaurice originally proposed). Secondly, Art. 60.2(c) allows any other party (not just the party specially affected by the breach) to suspend the treaty, in whole or in part, with respect to itself "if the treaty is of such a character that a material breach of its provisions by one party radically changes the position of every party with respect to the further performance of its obligations under the treaty".19 This provision resembles what Fitzmaurice wanted to see in respect of interdependent treaties (e.g., disarmament treaties).20 Thirdly and fourthly, the reference in Arts. 53 and 64 to "peremptory norms", conflict with which invalidates other treaties, is reminiscent of Fitzmaurice proposal to invalidate treaties in conflict with any treaty of an integral or interdependent nature. However, Arts. 53 and 64 do not cover all conflicts with integral treaties, only conflicts with integral treaties of a particular type, namely those of jus cogens. Fifthly and sixthly, Arts. 41 and 58 recalls Fitzmaurice proposal to invalidate inter se agreements in conflict with integral or interdependent treaties, when it outlaws (though not invalidates) inter se modifications to a multilateral treaty that "affect the enjoyment by the other parties of their rights under the treaty or the performance of their obligations" or relates to "a provision, derogation from which is incompatible with the effective execution of the object and purpose of the treaty as whole" (Art. 40.1(b)(i) and (ii)). Art. 58 provides for similar rules in respect of the inter se suspension of multilateral treaties.
Much aligned along Fitzmaurice's distinction between reciprocal and integral/interdependent obligations, James Crawford in his Third Report to the ILC on state responsibility distinguishes between "bilateral" obligations and "multilateral" obligations. In his view, "bilateral obligations can arise from a variety of sources, including general international law, bilateral or multilateral treaties or unilateral acts". Crawford, like Fitzmaurice, refers to the Vienna Convention on Diplomatic relations as an example of a multilateral treaty setting out legal relations that are essentially bilateral in character. As opposed to bilateral obligations, Crawford posits the notion of multilateral obligations. Such multilateral obligations are subdivided into two classes.
Firstly, multilateral obligations of the erga omnes type, "owed to the international community as a whole, with the consequence that all States in the world have a legal interest in the compliance with the obligation".21 This is the type of obligation erga omnes referred to by the ICJ in the Barcelona Traction case, itself distinguishing between reciprocal/bilateral obligations and integral/erga omnes obligations taking diplomatic relations as the standard example of the former type of obligations:
"an essential distinction should be drawn between the obligations of a State towards the international community as a whole, and those arising vis-à-vis another State in the field of diplomatic protection. By their very nature the former are the concern of all States".22
Erga omnes obligations are, in Crawford's view, "virtually coexistensive with peremptory obligations (arising under norms of jus cogens)".23
In the Final Draft Articles, adopted in August 200124, multilateral obligations of the erga omnes type are referred to as obligations "owed to the international community as a whole".25 The Commentary to the Final Draft Articles states that it is not its function to provide a list of those obligations. It refers instead to (i) the Barcelona Traction case which it gave the following examples: "the outlawing of acts of aggression, and of Genocide" and "the principles and rules concerning the basic rights of the human person, including protection from slavery and racial discrimination"26, and (ii) the East Timor case, where the ICJ added the right of self-determination of peoples to this list.27
The second class of multilateral obligations identified in the ILC's work on state responsibility are those owed erga omnes partes, i.e., owed, not to all states, but to all the parties to a particular regime (e.g., a multilateral treaty). In Crawford's view, this class concerns "obligations which are expressed (or necessarily implied) to relate to matters of the common interest of the parties". In other words, the performance of the obligations of each and every state party is recognised as being in the common interest of all state parties, common interest being defined as an interest "over and above any individual interest that may exist in a given case".28 As examples he refers to obligations that arise "in the fields of the environment (for example, in relation to biodiversity or global warming) and disarmament (for example, a regional nuclear free zone treaty or a test ban treaty)".29 Elsewhere Crawford rightly points out that "human rights obligations are not the only class of international obligations whose performance cannot be considered as affecting any `particular State' considered alone. This is also true of some obligations in such fields as human development, world heritage and environmental protection".30
In the Final Draft Articles, multilateral obligations of the erga omnes partes type are referred to as obligations "owed to a group of States ... and ... established for the protection of a collective interest of the group".31 Pursuant to the Commentary to the Final Draft Articles, multilateral obligations erga omnes partes
"must apply between a group of States and have been established in some collective interest. They might concern, for example, the environment or security of a region (e.g. a regional nuclear free zone treaty or a regional system for the protection of human rights). They are not limited to arrangements established only in the interest of the member States but would extend to agreements established by a group of States in some wider common interest".32
The Commentary explicitly states that it is not its function to provide an enumeration of collective interests giving rise to multilateral obligations erga omnes aprtes. It states, though, that "their principal purpose will be to foster a common interest, over and above any interests of the States concerned individually".33
The distinction between bilateral and multilateral obligations (in particular those of the erga omnes partes type) may not always be easily discerned. The Commentary to the Final Draft Articles states that "[i]t will be a matter for the interpretation and application of the primary rule to determine into which of the categories an obligation comes" and stresses that it only offers an "illustrative" discussion.34
The consequences attached to the distinction between bilateral and multilateral obligations in the Final Draft Articles relate to the question of standing.35 Legal standing to invoke responsibility for breach of a bilateral obligation is limited to the state at the other end of the bilateral relationship (that is, the "injured State").36 Breach of a multilateral obligation, in contrast, can be invoked either (i) by each and every one of the other parties to the multilateral treaty, in case of multilateral obligations erga omnes partes, or (ii) by any state, in case of obligations erga omnes, that is, binding on all states of the international community. For breach of multilateral obligations, two types of standing are, however, introduced. Only those states that are "specially affected" by the breach (so-called "injured States") are granted full standing, that is, standing to claim all of the remedies related to breach (cessation, non-repetition, reparation and countermeasures).37 All other, not "specially affected" (i.e., not "injured") states party to, or bound by, the multilateral obligation have standing only to claim cessation of the wrongful act and assurances and guarantees of non-repetition in the collective interest.38 They can also claim reparation, but only in the interest of the "injured State". The Commentary to the Final Draft Articles leaves it open as to whether these "non-injured" states can impose countermeasures.39 In any event, for those states to impose countermeasures, if at all possible, will be more difficult than for "injured States".
First of all, the problem of identifying whether an obligation is of a bilateral/reciprocal or multilateral/integral nature is not linked to the source of the obligation concerned. It is, in particular, not sufficient that an obligation derives from a multilateral treaty, for that obligation to be of the multilateral type (e.g., for all parties to that treaty to have standing to invoke breach of the treaty). The very question at issue is to distinguish between different types of obligations all of which derive from a multilateral treaty.40 Hence, the fact that WTO rules derive from a multilateral treaty is not enough for WTO obligations to be of the integral type.41
Secondly, the notion of multilateral/integral obligations should not be equated with obligations that are binding on states without their consent. Some integral obligations are binding on all states, irrespective of their consent, namely obligations deriving from norms of jus cogens.42 However, most integral obligations (such as those in the field of environmental protection) are binding only erga omnes partes, i.e., binding only on the states party to the treaty.
Thirdly, to say that obligations are reciprocal in nature does not necessarily mean that they are affecting only two governments in their bilateral relations. They may well affect also individuals or economic operators other than public authorities. The fact that obligations have a direct or indirect effect on individuals does not make them integral (although most integral obligations, being obligations "in the collective interest" will have a beneficial effect on individuals also). As clarified below, WTO rules are a perfect example of rules of a mainly reciprocal type that nonetheless have a clear effect on individual economic operators.
Fourthly, obligations of the reciprocal type should not be confused with obligations performance of which is inherently conditional on reciprocity. Although reciprocal in nature, reciprocal obligations may well be (and mostly are) unconditional, objective and self-existent in the sense that they must be complied with irrespective of compliance by other state parties (unless, of course, non-performance is justified as a suspension under Art. 60 or as a countermeasure). Here again, WTO rules are a perfect example. The WTO obligation to accord most-favoured-nation (MFN) treatment is an unconditional one: A WTO member must grant MFN status to imports of all other WTO members irrespective of how these other WTO members treat its own exports. MFN is, in that sense, not conditional on reciprocity. Nonetheless, although certain breaches of the MFN obligation may well affect all other WTO members, the MFN obligation is not as such an obligation of the multilateral/integral type. A breach of the MFN principle does not necessarily affect the rights of all WTO members (the discrimination may well be targeted at only one state). Rather, MFN is essentially a collection of equivalent bilateral legal relationships in respect of which no discrimination is allowed to take place. It is, in other words, an obligation of the bilateral/reciprocal type.
In Section A above, we set out the background and multiple legal consequences of the distinction between bilateral/reciprocal obligations and multilateral/integral obligations. It is time now to suggest a general criterion to make this distinction, knowing that the specifics of provisions dealing with particular consequences related to the distinction differ.
Reciprocal obligations are those where the legal matrix can be limited to a bilateral relationship between two states. Even if the obligation derives from a multilateral treaty binding on a large number of states, such reciprocal obligations can always be reduced to a compilation of bilateral state-to-state relationships. Hence, to affect or alter one of these bilateral relationships does not normally impunge on the other bilateral relationships, each of these relationships being detachable one from the other. Consequently, bilateral suspension or modification of reciprocal obligations is permissible and their breach can only be invoked by the state(s) at the other end of the bilateral relationship(s). The binding effect of integral obligations, on the other hand, is collective in nature. The different relationships between states bound by an integral obligation cannot be separated or detached into bilateral components. Hence, to affect or alter an integral norm will necessarily have an impact on all states bound by that norm. It cannot be limited to the bilateral relationship between two states. Consequently, bilateral suspension or modification of integral obligations is not permissible and their breach can be invoked by all states to which the obligation is owed.
In sum, the central question is the following: Does breach of the obligation in question necessarily affect the rights or obligations of all other states bound by the rule concerned? If so, the obligation is of the multilateral/integral type. If breach may affect the rights or obligations of only one, more or sometimes all other state parties, then it is of the bilateral/reciprocal type. This criterion shows that it is impossible to define a treaty in its entirety as reciprocal or integral in nature. One must look at every provision and every obligation individually.
Under this criterion, WTO obligations are of the reciprocal type. They are not integral in nature. A breach of WTO obligations does not necessarily affect the rights of all other WTO members. More than one, and in some instances all, WTO members may be so affected. But if all WTO Members are, indeed, affected this is not necessarily so, based on the nature of the WTO obligation. It is a consequence rather of WTO obligations being trade-related and trade restrictions, in turn, being capable of affecting the economic interests (not necessarily the rights) of many WTO members. That WTO obligations remain reciprocal and are not integral is, however, not always clearly discerned. The one author who, to my knowledge, examined this issue came to the same conclusion as I do here. Michael Hahn, after considering the hypothesis that GATT obligations are to be fulfilled erga omnes partes (and could, hence, be qualified as integral) reaches the conclusion that the basic structure of these obligations is against such qualification. He as well is of the view that both GATT and the WTO treaty remain treaties establishing bilateral right-obligation relationships between WTO members.43
Nonetheless, it must be recalled that the reciprocal versus integral distinction does not normally apply to treaties in their entirety. Indeed, even some rules in the WTO treaty are of the integral type, not because they are substantively more important than others, but for internal procedural/institutional reasons. The WTO integral rules in question are those relating to the operation of WTO bodies. When it comes to WTO rules setting out, for example, voting procedures, procedures on accession, the nomination of chairpersons or composition of certain WTO bodies, such rules must necessarily apply equally to all WTO members. The very nature of these procedural rules implies that their breach or inter se deviation would necessarily constitute breach towards all other WTO members (irrespective of trade or trade potential). As Schermers pointed out: "An organization can have only one constitutional structure. An amendment, for example, which expands the Executive Board of the organization from 18 to 24 members cannot be applied for some member States only".44 Hence, inter se modification of such institutional/procedural provisions cannot be tolerated, nor can state-to-state suspension of the obligations deriving from these provisions be legal as a form of countermeasures.45 However, the importance of these provisions being of the integral type are theoretical only (why would WTO members, inter se, want to change, for example, the election procedure of chairpersons?). What counts is that the substantive trade provisions in the WTO treaty are reciprocal in nature.
Having applied the general criterion to WTO obligations, we next elaborate on the main, underlying reasons why WTO obligations are reciprocal. We then attempt to explain why it is sometimes (mistakenly) thought that they are integral.
Firstly, and most importantly, trade is and remains a bilateral happening. Goods or services from one country are being exported or transferred to one other country.46 The rights and obligations negotiated in the WTO are aimed at ensuring market access for a given product from member A into the market of member B. In that sense, the WTO treaty is not all that different from the Vienna Convention on Diplomatic Relations (the standard example, referred to earlier, of a multilateral treaty imposing obligations of a bilateral/reciprocal nature): in the Vienna Convention, rights and obligations relate to diplomats sent from one country to another; the WTO treaty is about market access for goods from one country into another country. In the end, all WTO members may have similar market access rights (because of the MFN principle), but the rights thereby obtained remain trade-related, hence bilateral, in nature. Equally, a breach of WTO trade liberalisation obligations may have economic effects on more than one WTO member, because of the increased economic interdependence between states. But this is not the same as saying that a breach of WTO obligations necessarily affects the rights of all other WTO members, the way, for example, human rights or certain environmental law breaches do. A breach of WTO trade rules may affect a number of members individually, but it does not amount to an offence of the collective right or conscience of all state parties, the way a human rights breach does.
Trade (and hence WTO obligations) are international par excellence. A state cannot trade with itself. All WTO obligations relate, indeed, to foreign goods, foreign services or foreign service suppliers. Necessarily international, trade is also inherently bilateral. It takes the form of a physical or economic transfer of a unit from one country to another country. The benefits of WTO market access rights, as well as the welfare effects of compliance with WTO obligations, are spread over all WTO members (respectively, because of the MFN principle and a more efficient allocation of resources world-wide). But this collective effect does not negate the inherently bilateral character of trade and trade obligations.
The object and implementation of human rights and environmental treaties are, in contrast, a national matter. Especially respect for human rights is a matter between public authorities of a state and its own nationals. Steps taken to protect the environment are also national: laws and regulations are passed that set internal environmental standards within the territory of a particular state. Obviously, although human rights and environmental obligations are, in terms of object and implementation, a national or domestic matter, they are international and collective in terms of the values they protect and the effects they want to avoid. Respect for human rights has been elevated to the international level mainly because of the collective/universal values it seeks to achieve. Protection of the environment has been "internationalised" for reasons of effectiveness: environmental pollution knows no borders, hence for a country to effectively protect its environment it must co-operate with other countries.47 As a result, breach of environmental and, especially, human rights obligations becomes of international interest, but this interest is essentially collective. Quite often (as will be the case of, for example, violations of the Kyoto Protocol) not one particular state will be more affected than another. The interest protected is collective, not bilateral.
In sum, although trade is inherently international its obligations are bilateral; whereas human rights and environmental protection are inherently national, most of its obligations are collective.
Secondly, unlike, for example, the prohibition of genocide or the protection of human rights or the environment, trade and the liberalisation of trade is not a value. It is not sought after for the achievement of some "global common" that transcends the sum total of individual state interests. Trade is not a value, it is only an instrument. It is an instrument to increase the economic welfare of all states. But the increase in welfare thus created by trade does not exceed the sum of economic welfare experienced by its composite members. Of course, it can be expected that higher economic welfare will translate itself also into social and other benefits, but this would be the result mainly of what states or private operators themselves decide to do with their welfare, not a direct consequence of trade or WTO rules.
Trade and WTO provisions seeking to enhance trade are, therefore, not like, for example, the Genocide Convention where, in the words of the ICJ: "the contracting States do not have any interests of their own; they merely have, one and all, a common interest, namely, the accomplishment of those high purposes which are the raison d'être of the convention. Consequently, in a convention of this type one cannot speak of individual advantages to States, or of the maintenance of a perfect contractual balance between rights and duties".48 The same has been said in respect of human rights treaties. As the Inter-American Court of Human Rights pointed out: "human rights treaties ... `are not multilateral treaties of the traditional type concluded to accomplish the reciprocal exchange of rights for the mutual benefit of the contracting States'; rather `their object and purpose is the protection of the basic rights of individual human beings, irrespective of their nationality, both against the State of their nationality and all other contracting States'".49 Or as the ECHR noted: "the purpose of the High Contracting Parties in concluding the Convention was not to concede to each other reciprocal rights and obligations in pursuance of their individual national interests but to realise the aims and ideals of the Council of Europe, as expressed in its Statute, and to establish a common public order ... [I]t follows that the obligations undertaken by the High Contracting Parties in the Convention are essentially of an objective character, being designed rather to protect the fundamental rights of individual human beings from infringement by any of the High Contracting Parties than to create subjective and reciprocal rights for the High Contracting Parties themselves".50
The same objective or "integral" nature has been attributed to EC treaties. This was done not so much because of the "higher values" protected by EC treaties, but because EC law was construed as setting up a "new legal order" that confers rights on individuals and is supreme to national law.51 In addition, under EC law elements of exclusive competence have been granted to EC institutions (e.g., to regulate areas such as external trade).52 In that sense, EC treaties have become a form of domestic law, in respect of which member states (much like individuals in internal law) have no competence to deviate from inter se. To that extent, obligations under EC law can, indeed, be characterised as integral in nature.
WTO obligations, in contrast, give states an instrument to achieve other "nobler" goals. Since, therefore, the WTO treaty is only an instrument to achieve certain other goals, agreements modifying the WTO treaty inter se in pursuit of these other goals must, generally speaking, be accepted as long as they do not affect the rights of third parties.53 This is one of the major consequences of characterising WTO obligations as bilateral/reciprocal in nature.
Thirdly, the fact that WTO obligations are reciprocal in nature can be deduced from the way they are negotiated and re-negotiated and, in particular, with reference as to how they are enforced.
WTO obligations, especially those set out in country-specific schedules of concessions, were negotiated first state-to-state, on a bilateral level: state A gives and takes; state B does the same. This bilateral and mutual reduction in trade restrictions is then multilateralised and applied, respectively, by state A and state B in their bilateral relationships with all other WTO members. As the Appellate Body remarked: "Tariff negotiations are a process of reciprocal demands and concessions, of `give and take'".54 The ultimate aim of this "give and take" exercise is to achieve an appropriate balance of trade concessions. Or as the third preamble to the Marrakesh Agreement put it, the underlying objectives of the WTO are to be achieved by "entering into reciprocal and mutually advantageous arrangements".
Also the way GATT and GATS concessions are re-negotiated under GATT Art. XXVIII and GATS Art. XXI proves the reciprocal nature of WTO obligations. Essentially, only other WTO members with a substantial trade interest must be involved and agree to the re-negotiated list of concessions. Re-negotiation, once again, happens largely on a bilateral level.
Crucially, however, the taste of the pudding comes with the eating: the way WTO obligations are enforced is exclusively bilateral. WTO dispute settlement does not, in the first place, tackle breach, but rather nullification of benefits that accrue to a particular member.55 Panel and Appellate Body proceedings only examine claims made by one WTO member against one other WTO member. Most importantly, in case the defendant loses and does not comply within a reasonable period of time, the winning state will be authorised to impose state-to-state countermeasures against the losing state (DSU Art. 22). This exclusively bilateral modality of enforcement of WTO rules is an important indication that most WTO obligations are reciprocal in nature. Especially the fact that the WTO treaty allows one member to suspend its WTO obligations as a form of countermeasure towards one other member proves that WTO obligations are not of the integral type. If WTO obligations were of the integral type, their inter se suspension would necessarily affect the rights of all other WTO members and thus fall afoul of the pacta tertiis principle (confirmed in Art. 58 of the Vienna Convention) as well as the general rule that countermeasures may suspend only rights of the violating state, not of any third state.56 Although third states may well feel an economic effect of the inter se suspension (an effect that may be positive or negative), their WTO rights will not normally be affected. In respect of integral obligations the situation is different: their inter se suspension or modification cannot be tolerated. The obligation, in those cases, is of an "integral" type in that it ought to be respected in all circumstances (unless, of course, exceptions are provided for), irrespective of the conduct of other states.57 This explains, for example, why fundamental human rights obligations (which are obligations of the integral type) cannot be suspended in response to breach58: their suspension towards the wrongdoing state would not only affect that state, but also breach the rights of all other contracting parties.
Of course, the general bilateral mode of enforcement of WTO rules cannot as such be seen as conclusive proof that all WTO obligations are reciprocal in nature. Human rights and environmental obligations (most of which are accepted as being of an integral nature) can also be enforced on a purely bilateral, state-to-state basis (although under these regimes a collective non-compliance mechanism is more common). What such integral obligations would, nonetheless, not allow for is the suspension of obligations as a form of retaliation, the way WTO suspension works. In that sense, WTO suspension as a form of countermeasures is, in and of itself, sufficient proof that WTO obligations are not integral in nature. At the same time, although WTO obligations are reciprocal in nature, nothing prevents WTO members from setting up a collective non-compliance mechanism along the lines of certain environmental treaties (to some extent, the trade policy review mechanism does exactly that). Such collective non-compliance mechanism would not alter the character of WTO obligations and somehow transform them into integral obligations simply because they are enforced collectively. The efficient implementation of WTO rules may, indeed, be well, if not better, served by a collective compliance mechanism, instead of the current bilateral, state-to-state litigation system centred around breach and bilateral countermeasures.59
A first confusion that could arise is to take the increase in global welfare that trade liberalisation (and thus most WTO obligations) brings about, as evidence that WTO obligations are a "global common", in the collective interest, and, for that reason, must be integral in nature. Trade liberalisation is, indeed, in the general interest in that it should increase overall global welfare as a result of a better allocation of the world's resources, including the welfare of the state making a particular trade "concession". Much like environmental obligations of a mostly integral nature, WTO rules are, therefore, in the "general interest". However, unlike environmental obligations, the interest achieved by WTO obligations remains a compilation of individual welfare increases, not the achievement of a "global common", such as the preservation of the planet's climate system, which transcends the individual benefits of individual states. The fact that engaging in an obligation is to everyone's individual benefit, including the one engaging in the obligation does not mean that it is in the collective interest in the sense of "a common interest, over and above any interests of the States concerned individually".60
After all, the fact that the WTO treaty, with its trade liberalising obligations, is in the interest of all WTO members may not be that special. Is not every international treaty concluded by states supposed to be in the mutual interest of these states? This point is best illustrated with reference to the Vienna Convention on Diplomatic Relations, generally accepted as setting out reciprocal obligations. There as well, the fact that the provisions of this Convention are undoubtedly in the "general interest" of all states (as confirmed by the ICJ in the Diplomatic and Consular Staff case61), does not alter the bilateral/reciprocal nature of the obligations they set out.
Finally, to state that trade liberalisation is, generally
speaking, beneficial to all WTO members is one thing, to say that each and
every WTO rule is of this "public interest" nature, quite another. As
Benedek pointed out: "the relevance of the `legal economics' in international
economic law is questioned in view of the often strong political element
involved in international economic relations which makes it unrealistic to
regard the GATT rules from the perspective of a kind of `economic law of
nature'".62 Witness, for
example, WTO rules on anti-dumping which for many do not make economic sense
and the GATT/WTO's reluctant incorporation of the agricultural and textiles
sector into mainstream trade liberalisation rules, both as a result of strong
political lobbying by developed country interests.
The multiple MFN obligations set out in the WTO treaty, the cornerstone of the multilateral trading system, ensure that any trade advantage a country gives to another must be "multilateralised" and granted to all WTO members. As a result, MFN surely makes bilateral concessions collective in the sense that they must go to all other WTO members. But in substance, this "collectivisation" is nothing more than a duplication by the number of WTO members of the original bilateral concession. The bilateral concession is thereby granted from state to state to all other WTO members. It does not, by means of MFN, transcend into some "global common", more valuable than the sum total of the individual benefits it procures to each WTO member. Looked at from a different angle, MFN is, of course, an obligation owed towards all WTO members, but when member A discriminates only member B (e.g., by banning all imports coming from B), this MFN breach can hardly be said to affect the MFN right of members C, D and E (who can continue to export to member A, arguably even more so than before given that the ban has stopped the supply coming from member B).
Finally, given that (i) compliance with WTO rules normally achieves an increase in welfare world-wide and (ii) the economic inter-dependence between states is ever increasing, a breach of WTO obligations by one member is likely to affect, directly or indirectly, the economic interests of many, sometimes all, other WTO members. This may wrongly be interpreted as granting a form of actio popularis to all WTO members for each and every breach of WTO rules, irrespective of the states involved. In turn, given that all WTO members could then (under this wrong assumption) complain about any WTO breach, this element could be mistakenly seen as proof that WTO obligations are of an erga omnes partes or integral nature.
In EC - Bananas the Appellate Body decided that the United States could bring a case under GATT even though it hardly produces bananas and has not yet exported any. The Appellate Body quoted with approval the following remark from the panel report:
"with the increased interdependence of the global economy ... members have a greater stake in enforcing WTO rules than in the past since any deviation from the negotiated balance of rights and obligations is more likely than ever to affect them, directly or indirectly".63
One must, however, distinguish three issues: firstly, when does a WTO member have legal standing to bring a WTO complaint (the matter at issue in EC - Bananas); secondly, what is required for a WTO complaint to be validly established; thirdly, what are the consequences of validating a WTO complaint? We examine these three distinct issues in turn.
On the first issue of standing, the Appellate Body in EC - Bananas made the rather astonishing finding that in order to bring a case under the DSU, no "legal interest" is required. In particular, it did "not read any of [the PCIJ/ICJ] judgements [referred to by the EC] as establishing a general rule that in all international litigation, a complaining party must have a `legal interest' in order to bring a case".64
Looking at the cited international case law65 and the ILC Final Draft Articles, it may be so that not much reference is made there to "legal interest" (in the sense of an interest to see the law abided by), but this is so because normally more than a "legal interest" is needed for a state to have standing, namely one must prove the existence of a "legal right". Any breach of international law could be said to affect the legal interest of all other states, that is, the interest of states to see respect for the rule of law in general (even if these states draw no individual rights from this law).66 But this "legal interest" is not normally enough for a state to invoke responsibility for breach.67 Even if Art. 48 of the Final ILC Draft Articles grants certain rights to enforce state responsibility to states that are not "injured States"68, but merely have a legal interest, this occurs only for international norms of a certain nature, e.g., obligations erga omnes or erga omnes partes; not for obligations of a reciprocal nature, such as most WTO obligations. As Crawford noted in his Third Report:
"outside the field of 'integral' obligations, or obligations erga omnes partes, ... it is doubtful that States have a right or even a legally protected interest, for the purposes of State responsibility, in the legal relations of third States inter se".69
Hence, even for those breaches of international law in respect of which the most lenient rules on standing apply (say, erga omnes or jus cogens obligations), standing is granted because of the existence of a "legal interest". Hence, in all cases where standing is acknowledged, at least a "legal interest" must be pointed at. Therefore, the only way to make sense of the Appellate Body finding that no "legal interest" is required, is to assume that the Appellate Body did not mean to refer to "legal interest" in its usual sense of "interest to see the law abided by", but understood this term to mean, for example, a requirement of proof of actual damage or trade diversion.
It is important to recall what the Appellate Body did not state in EC - Bananas. It did not say that a purely "legal interest" to see WTO rules abided by is sufficient for any WTO member to have standing in respect of all possible breaches of WTO law. On the contrary, it stated that there is no requirement of "legal interest". Of course, like any WTO member (and arguably even non-WTO members) the United States did have a "legal interest" to see GATT rules abided by. In addition, however, the Appellate Body was careful enough to base its conclusion that the United States did have standing under GATT on other factors as well (not related to purely "legal interest"): the United States was a producer of bananas and hence a potential exporter, the US market for bananas was potentially affected by the EC regime in terms of world supplies and prices and the GATT claims were inextricably interwoven with those under GATS for which the United States did unmistakably have standing.70 The Appellate Body stressed that
"taken together, these reasons are sufficient justification ... This does not mean, though, that one or more of the factors ... would necessarily be dispositive in another case".71
The Appellate Body forgot, however, to mention one other obvious but crucially important factor for standing to be granted, namely the fact that the EC import regime for bananas, if found to be discriminatory the way the United States claimed it to be, was favouring ACP countries and hence, at least in theory, discriminating all other WTO members, including the United States. The measure at issue was not a measure discriminating only one other WTO member (e.g., a member other than the United States), nor a trade restriction that did, on the books, not apply to US exports, the way, for example, an allegedly WTO inconsistent anti-dumping duty imposed by the EC on cement from Mexico would have nothing to do with US rights under the WTO.
In sum, based on a close reading of the Appellate Body decision in EC - Bananas, the mere fact that a WTO member breaches WTO rules does not suffice for all other WTO members to have standing to seek redress for this breach. A purely "legal interest" is not enough.72 For a member to have standing the inconsistent measure must, at least in theory, apply to the trade of that member. An anti-dumping duty imposed by the United States on steel from Japan only cannot be challenged by the EC; nor can the United States complain about Zambia discriminating exports from Nigeria only. In addition, even if the measure does de jure apply to the trade from that other member, some proof must be provided that either actual or potential trade flows are being restricted and/or that the member is otherwise economically affected (e.g., by an increase in world prices, as referred to in EC - Bananas). The first condition will, in effect, constitute a bigger hurdle than the second. Indeed, if the measure does, on the books, apply to the complainant, trade potential or missed trade opportunities will be easily proven, in particular given the economic inter-dependence of WTO members.73
The fact that these two conditions for standing do exist, and that, therefore, the WTO does not know an actio popularis, shows that WTO obligations are reciprocal in nature, not integral or binding erga omnes partes. Hence, absence provisions in the WTO treaty to the contrary, the general rule in Art. 42 (a) of the Final ILC Draft Articles applies.74 That is, a WTO member can only bring a complaint against another WTO member in case the obligation allegedly breached is owed to it "individually".
Here again, the traditional example of reciprocal/bilateral obligations, namely those in the Vienna Convention on Diplomatic Relations, is instructive. As is the case for breach of WTO obligations, breach of obligations under the Vienna Convention is mostly likely to affect also other state parties.75 However, this does not necessarily give those other states standing to invoke responsibility for the breach, nor does it detract from the bilateral/reciprocal nature of Vienna Convention obligations. Only in case the breach is one of an obligation owed to the state "individually" -- say, only in case the alleged wrongdoer has violated diplomatic law as against diplomats of the complainant state -- will the complainant state have standing.
As pointed out, nothing prevents states, when concluding a bilateral/reciprocal treaty like the WTO treaty, from "contracting out" of those general international law rules on standing. Hence, WTO members (or, for that matter, the Appellate Body) could well decide that, for whatever policy reason, it is desirable to make breach of WTO law challengeable by all WTO members, irrespective of the breach. Such would not, in and of itself, change the nature of WTO obligations as reciprocal obligations. (In contrast, the fact that, as the law now stands, not every WTO member can complain about breach of WTO law, does prove that WTO obligations are reciprocal in nature, not binding erga omnes partes). The main argument against giving standing to all WTO members, based on a general legal interest to see the treaty abided by, is the risk of effectively appointing a number of powerful states as public prosecutors or policemen, with the result that especially (or only) obligations in the particular national interest of those states would be enforced. In the WTO, granting such actio popularis could mean that WTO agreements on, say, intellectual property or trade in services are more often judicially enforced than, for example, the agreements on Agriculture or Textiles and Clothing.76
The second condition for a member to have standing (proof of trade effects or, at least, trade potential or opportunities) is directly related to the conditions for a breach of WTO obligations to be established. Since standing essentially depends on the existence of a legal right, proof of a legal right will be inextricably linked to proof of breach. Now, for a breach of WTO rules to be established, it is generally accepted in GATT/WTO case law that a complainant is not required to prove that actual trade flows have been diverted.77 Proof of trade opportunities being effected will be enough. This explains, at the same time, why prove of trade effects is not a requirement either in terms of standing.
Nonetheless, as pointed out earlier, for a complainant to prove that a measure violates WTO rules is not enough for that measure to be condemned by a panel. In addition, the complainant must point at nullification or impairment of benefits accruing to it.78 As noted above, this element supports the view that WTO obligations are reciprocal in nature. However, DSU Art. 3.8 provides for a presumption to the effect that breach "is considered prima facie to constitute a case of nullification and impairment". This is further explained to mean that "there is normally a presumption that a breach of the rules has an adverse impact on other Members". However, as the Appellate Body noted, Art. 3.8 is about "what happens after a violation is established".79 It does not relate to the issue of standing required in order to be allowed to invoke breach. In other words, to say that nullification of benefits is presumed once breach is established (as Art. 3.8 does) is not the same as saying that all WTO members have a right to complain about any WTO breach.
Even so, the second criterion referred to above for a member to have standing (proof of at least trade potential or opportunities) must run parallel with the requirement for there to be nullification. In EC - Bananas, for example, the Appellate Body referred to the same factors under both standing and nullification under DSU Art. 3.8, namely the United States being a producer and potential exporter and the possible effects on the US internal market. It noted that "[t]hese matters that we have already decided are relevant to the question of the standing of the United States under the GATT 1994. They are equally relevant to the question whether the European Communities has rebutted the presumption of nullification or impairment".80
Consequently, once a WTO member has been granted standing, in particular, once it has met the second criterion for it to have standing, it must necessarily meet also the nullification and impairment condition in DSU Art. 3.8. As a result, once standing accorded and once breach established, the presumption of nullification must, for all practical purposes, be irrebutable.
Although breach of WTO rules may be successfully invoked by a number of WTO members (meeting, first, the two rather lenient conditions for standing and, second, the relatively wide definition of breach, both referring to trade opportunities, not trade effects), a finding of breach of WTO rules is a purely bilateral matter. Quite often, whether or not there is breach will actually be dependent on the complainant (are its exports, for example, "like products" as compared to the domestic products allegedly protected in breach of, say, GATT Art. III:4?). A finding of WTO breach is bilateral also in the sense that only the complainant, not other WTO members, may directly rely on it. Only the complainant may suspend concessions if no implementation follows, not other WTO members. In addition, the fact that other WTO members have exactly the same measure in place as the one found to be in breach does not give other WTO members any rights to suspend concessions. A new state-to-state complaint must be lodged against these equivalent measures for their WTO illegality to be conclusively established.
These bilateral consequences of breach, combined with the requirements for standing set out above and the state-to-state definition of breach (including the condition of member-specific nullification), are proof that WTO obligations are not integral, but reciprocal.
The fact that WTO obligations are bilateral/reciprocal obligations (not multilateral/integral obligations) puts them in perspective. It stresses their relative importance as obligations of an essentially contractual type that can, in principle, be deviated from or suspended inter se (in line with Arts. 41 and 58 of the VCLT).81 They are not integral in the sense of being immutable obligations to be respected at all times and as between all WTO members, irrespective of other norms of international law. WTO obligations are, in this sense, framework obligations only or lex generalis that can be supplemented or deviated from, as between some or all WTO members, by other rules of international law (especially, rules that are of an integral type).
WTO law is but a branch of public international law. Given its reciprocal nature, WTO law will, moreover, have to allow for and give way to a number of other rules of international law. The fact that some WTO members agree to deviate from WTO obligations in their inter se relations only (e.g., by mutually agreeing to condition their trade on respect for human rights) must, in principle, be accepted, given the reciprocal nature of WTO obligations. As long as the rights of other WTO members are not affected, these other WTO members, not party to the inter se deviation, cannot complain about these purely inter se contractual changes. The requirements for other members to have standing to challenge such inter se deviations will then, indeed, not be met (on the books, the measure does not even apply to their trade, not even their potential trade). If WTO obligations were, however, of the integral type, no such inter se modifications could have been tolerated.
Applying these conclusions to, for example, conflicts between WTO law (of a bilateral nature) and provisions in multilateral environmental agreements (MEAs) or human rights treaties (mostly of a multilateral/integral nature), the following solutions can be offered.
In case the WTO rule can be defined as the later in time, it can be seen as an inter se agreement modifying certain integral MEA or human rights obligations. If this is the case -- i.e., if the WTO rule deviates from integral obligations as between WTO members only -- then WTO rules would not only affect WTO members but also third party rights (i.e., rights of non-WTO members, party to the human rights treaty or MEA) or even be incompatible with "the effective execution of the object and purpose of the [human rights or MEA] treaty as a whole" (in the sense of Art. 41, para. 1(b)(ii)). Consequently, as between parties to the MEA or human rights treaty, the particular WTO provision, to the extent of the conflict, would then be illegal under Art. 41 VCLT. In other words, the conflict should then be decided in favour of the earlier MEA or human rights provision. This would be the result of Art. 41 VCLT. The fact that the WTO rule is the lex posterior under Art. 30.4(a) VCLT does not alter this solution. Art. 30.5 VCLT explicitly makes the operation of Art. 30.4 subject to Art. 41.82
In contrast, if the human rights treaty or MEA would be the later in time, it must be recalled that inter se agreements modifying reciprocal obligations such as those in the WTO treaty are, in principle, accepted (pursuant to Art. 41 VCLT). The human rights treaty or MEA could then be defined as an inter se agreement deviating from WTO rules as between the parties to the human rights treaty or MEA. Given the reciprocal nature of WTO obligations, this inter se agreement is unlikely to be illegal under Art. 41 (unless it affects the rights of third parties). Hence, in the event of conflict, the later human rights treaty or MEA would not be prohibited as an illegal inter se deviation. It should then prevail as the lex posterior pursuant to Art. 30.4(a) (of course, only as between the parties to both norms). The earlier WTO rule is then, however, not illegal. It is simply the lex prior having to give way inter se to the later human rights treaty or MEA.
In sum, irrespective of the actual timing of the two norms, in the event of conflict between a WTO rule of the reciprocal type and a human rights or MEA obligation of an integral nature, the human rights or MEA obligation must, either pursuant to Art. 41 VCLT or pursuant to Art. 30.4(a) VCLT, prevail in the relationship between two parties that are bound by both norms.
Take the example of conflict between a WTO obligation not to restrict trade in product X (assuming that GATT Art. XX does not allow for trade restrictions on product X) and an MEA or human rights obligation, of the integral type, to restrict trade in product X because product X is defined as an environmentally or socially unfriendly good. As between states that are bound by both rules, the MEA/human rights rule must prevail, irrespective of whether it comes earlier or later in time. If it comes later in time, it prevails as the lex posterior under Art. 30.4(a) VCLT. If it is the earlier in time, it cannot, as an integral obligation, be deviated from inter se, by the later WTO rule.
1 ICJ Reports 1951, 23.
2 Ibid. In his Dissenting Opinion, Judge Alvarez went even further, classifying treaties of the type of the Genocide Convention as follows: "To begin with, they have a universal character; they are, in a sense, the Constitution of international society, the new international constitutional law. They are not established for the benefit of private interests but for that of the general interest" (ibid., 51, emphasis in the original).
3 Ibid., 24.
4 PCIJ, Series A/B, No. 41, 64 (1931).
5 PCIJ, Series A/B, No. 63, 132-134 (1934). For other cases where a treaty was characterised as transcending the interests of the parties directly concerned and as constituting a so-called objective regime, binding even on non-parties, see the Wimbledon case, where the PCIJ found that the international regime for the Kiel Canal (set out in the Versailles Peace Treaty) was binding also on Germany, even though Germany was not a party to the treaty (PCIJ Reports 1923, Series A, No. 1) and the Dispute on the Regime of Demilitarization for the Aaland Islands, where an ad hoc Committee of Jurists decided that the Paris peace settlement of 1856 setting out international obligations on demilitarisation was binding also on, and could be invoked by, Sweden and Finland, even though they were not a party to the settlement (see Report of the International Committee of Jurists entrusted by the Council of the League of Nations with the task of giving an advisory opinion upon the legal aspects of the Aaland Islands Question, League of Nations Official Journal, Special Supplement No. 3, October 1920).
6 Third Report on the Law of Treaties by Gerald Fitzmaurice, UN doc. A/CN.4/115, YILC, Vol. II, 20 (hereafter "Fitzmaurice, Third Report"), 27, Art. 18, para. 2.
7 Ibid., 27, Art. 19.
8 Second Report on the Law of Treaties by Gerald Fitzmaurice, UN doc. A/CN.4/107, YILC, Vol. II, 16 (hereafter, "Fitzmaurcie, Second Report"), 54.
9 Ibid., 55.
10 Ibid., Art. 19.
11 Fitzmaurice, Third Report, Art. 18.
12 Fitzmaurice, Second Report, Art. 19(iv).
13 Fitzmaurice, Third Report, Art. 19.
14 Fitzmaurice, Second Report, Art. 29.1(iii).
15 Ibid., Art. 19.1(ii)(b). Or as he noted in respect of the example of disarmament treaties: "the obligation of each party to disarm ... is necessarily dependent on a corresponding performance of the same thing by all the other parties, since it is of the essence of such a treaty that the undertaking of each party is given in return for a similar undertaking by the others" (Ibid., 54).
16 Fitzmaurice, Third Report, Art. 19.
17 James Crawford, in his Third Report on State Responsibility (UN doc. A/CN.4/507, 10 March 2000, paras. 99-108, hereafter "Crawford, Third Report") talks about reciprocal and integral obligations, not about interdependent ones.
18 In the same vein, the ILC's Final Draft Articles on State Responsibility (Art. 50) prohibit the taking of countermeasures affecting: « (a) The obligation to refrain from the threat or use of force as embodied in the Charter of the United Nations ; (b) Obligations for the protection of fundamental human rights ; (c) Obligations of a humanitarian character prohibiting reprisals ; (d) Other obligations under peremptory norms of general international law ».
19 For an analogy in the law on state responsibility, see supra note 37.
20 In contrast, for integral treaties, Fitzmaurice wanted to outlaw any termination or suspension. Hence, it is incorrect for the ILC in its commentary to Art. 40 (para. (19)) of the 1996 Draft Articles (Draft Articles on State Responsibility adopted by the ILC on First reading, ILC 48nd session, 1996) and James Crawford in his Third Report (at para. 91) to refer to Art. 60.2(c) as an expression of Fitzmaurice's theory on "integral obligations". The consequences in Art. 60.2(c) are rather those Fitzmaurice wanted to see in respect of "interdependent" treaties.
21 Crawford, Third Report, para. 106 (a).
22 Barcelona Traction case (Second Phase), I.C.J. Reports 1970 at pp. 32-3 (paras. 33-34). For other ICJ pronouncements in respect of erga omnes obligations, see Namibia Opinion, I.C.J. Reports 1971 p. 16 at p. 56 (para. 126); Case concerning East Timor, I.C.J. Reports 1995 p. 90 at p. 102 (para. 29); and Application of the Convention on the Prevention and Punishment of the Crime of Genocide. Bosnia and Herzegovina v. Yugoslavia (Preliminary Objections), I.C.J. Reports 1996 p. 625 at p. 626 (para. 4), p. 628 (para. 6).
23 Crawford, Third Report, para. 106 (a).
24 ILC Report on the work of its fifty-third session (23 April - 1 June and 2 July - 10 August 2001), General Assembly, Official Records, Fifty-fifth Session, Supplement No. 10 (A/56/10), posted at « http://www.un.org/law/ilc/reports/2001/2001report.htm ». The Draft Articles are on pp. 43-59 [hereafter « Final Draft Articles »], the Commentary thereto on pp. 59-365 [hereafter « Commentary »].
25 Final Draft Articles, Art. 48, para. 1(b).
26 Barcelona Traction, Light and Power Company, Limited, Second Phase, ICJ Reports 1970, p. 3, at p. 32, para. 34.
27 ICJ Reports 1995, p. 90, at p. 102, para. 29.
28 Crawford, Third Report, para. 92.
29 Ibid., para. 106 (b). In note 195, Crawford submits that « integral obligations » are a sub-category of obligations erga omnes partes. As noted earlier (supra note 20), he seems to be incorrect when referring to the treaties mentioned in Art. 60(2)(c) of the Vienna Convention as "integral treaties". Rather, they are what Fitzmaurice called "interdependent treaties". The way Fitzmaurice saw "integral treaties" should classify them rather across both obligations erga omnes (Fitzmaurice's prime example of an integral treaty was, after all, the Genocide Convention) and obligations erga omnes partes (such as most human rights and environmental provisions).
30 Ibid., para. 88. On that basis, Crawford rightly criticises the ILC Draft 1996 for singling out "human rights and fundamental freedoms" in its Art. 40.2(e)(iii).
31 Final Draft Articles, Art. 48, para. 1(a).
32 Commentary, p. 320-1, para. (7).
34 Commentary, p. 297, para. (6).
35 For related consequences see Art. 50, quoted supra note 18, prohibiting countermeasures that affect certain multilateral obligations.
36 Art. 42 (a) of the Final Draft Articles provides: "A State is entitled as an injured State to invoke the responsibility of another State if the obligation breached is owed to: (a) That State individually ...".
37 Art. 42 (b) states: "A State is entitled as an injured State to invoke the responsibility of another State if the obligation breached is owed to ... (b) A group of States including that State, or the international community as a whole, and the breach of the obligation: (i) Specially affects that State; or (ii) Is of such a character as radically to change the position of all the other States to which the obligation is owed with respect to the further performance of the obligation". The obligations referred to in Art. 42 (b)(ii) are those that Fitzmaurice termed "interdependent" obligations, discussed earlier. The Commentary to the Final Draft Articles (at p. 300, para. (13)) defines them as "obligations, breach of which must be considered as affecting per se every other State to which the obligation is owed" and makes the analogy with Art. 60 (2) (c) of the VCLT. According to the Commentary "[e]xamples include a disarmament treaty, a nuclear free zone treaty, or any other treaty where each parties performance is effectively conditioned upon and requires the performance of each of the others" (ibid.). It is stated to be "desirable that this subparagraph be narrow in its scope" (p. 301, para. (15)). Like James Crawford (see supra notes 20 and 29), the Commentary wrongly equates "integral" to "interdependent" obligations on p. 296, para. (5) and in footnote 706.
38 Art. 48, para. 2.
39 At para. (8) on pp. 327-8, it is stated: "The focus of the chapter is on countermeasures taken by injured States as defined in article 42. Occasions have arisen in practice of countermeasures being taken by other States, in particular those identified in article 48, where no State is injured or else on behalf of and at the request of an injured State. Such cases are controversial and the practice is embryonic. This chapter does not purport to regulate the taking of countermeasures by States other than the injured State."
40 Moreover, technically speaking also an obligation deriving from a bilateral treaty could be of the integral type. Imagine, for example, that at the origin of the 1948 Genocide Convention, two states had first concluded a bilateral treaty outlawing genocide.
41 See the Commentary to Art. 42 (a) of the ILC Final Draft Articles (at p. 298, para. (8)), confirming that the "bilateral obligations" referred to there are "intended to cover cases where the performance of an obligation under a multilateral treaty or customary international law is owed to one particular State ... although a multilateral treaty will characteristically establish a framework of rules applicable to all the States parties, in certain cases its performance in a given situation involves a relationship of a bilateral character between two parties. Multilateral treaties of this kind have often been referred to as giving rise to `bundles of bilateral relations'".
42 See also supra note 5.
43 MICHAEL HAHN, DIE EINSEITIGE AUSSETZUNG VON GATT-VERPFLICHTUNGEN ALS REPRESSALIE 396 (as well as Part 3) (1995).
44 Schermers, The Legal Basis of International Organization Action, in A HANDBOOK ON INTERNATIONAL ORGANIZATIONS (ED. RENE-JEAN DUPUY), 401-411, at 409 (1998).
45 Art. X:2 of the Marrakesh Agreement confirms this in respect of Art. IX on "decision-making", providing that amendments to Art. IX "shall take effect only upon acceptance by all Members".
46 Several countries may, of course, have been involved in the production of a particular good, but rules of origin are exactly there to determine the origin of each and every particular good. The fact that a good can, legally speaking, originate only in one country confirms the bilateral nature of trade.
47 Often a triple distinction is made between multilateral environmental agreements: they may seek to regulate trade in a particular category of products (such as wildlife), to protect states from substances harmful to their domestic environment (such as hazardous waste) or to protect so-called global commons such as the ozone layer or the global climate system. All three types require cooperation across borders. Especially the third type must be classified as being of an integral nature (protecting global commons).
48 ICJ Report 1951, 23.
49 Inter-American Court of Human Rights, Advisory Opinion of 8 September 1983, quoted in Rudolf Bernhardt, Thoughts on the Interpretation of Human-Rights Treaties in PROTECTING HUMAN RIGHTS: THE EUROPEN DIMENSION, STUDIES IN HONOUR OF G. J. WIARDA (ED. F. MATSCHER AND H. PETZOLD), 65-71, at 68-69.
50 Case 788/60, 4 YEARBOOK OF THE EUROPEAN CONVENTION ON HUMAN RIGHTS 138, at 140 (1961).
51 See Case 26/62 Van Gend en Loos  ECR 1. The importance of the new legal order in separating the EC from other international organisations was reiterated in the ECJ's Opinion on the draft Agreement on a European Economic Area, Opinion 1/91,  ECR I-6079,  I CMLR 245.
52 See Case 6/64  ECR 585, at pp. 593-4 (Costa v. ENEL): "By creating a Community of unlimited duration, having its own institutions, its own personality, its own legal capacity and capacity of representation on the international plane and, more particularly, real powers stemming from a limitation of sovereignty or a transfer of powers from the States to the Community, the Member States have limited their sovereign rights, albeit within limited fields, and have thus created a body of law which binds both their nationals and themselves. The integration into the laws of each Member State of provisions which derive from the Community, and more generally the terms and the spirit of the Treaty, make it impossible as a corollary, to accord precedence to a unilateral and subsequent measure over a legal system accepted by them on a basis of reciprocity".
53 See Arts. 41/58 of the VCLT.
54 Appellate Body report on European Communities - Customs Classification of Certain Computer Equipment, complaint by the United States (WT/DS62/AB/R), adopted on 22 June 1998, para. 109.
55 Even if under the DSU such nullification is presumed once breach is established (DSU Art. 3.8), see infra.
56 See Art. 49.1 of the ILC Final Draft Articles ("An injured State may only take countermeasures against a State which is responsible for an internationally wrongful act ...").
57 Recall, however, the special situation of so-called "interdependent" obligations where breach or suspension by one state party may well make compliance by the other states impossible and allow all of the other state parties to suspend the treaty. The classical example is a disarmament treaty. See Art. 60, para. 2(c) VCLT and Art. 42 (b)(ii) of the Final Draft Articles.
58 See Art. 50, para. 1(b) of the Final Draft Articles, quoted surpa in note 18. See also Art. 60.5 of the VCLT.
59 See M. Pinto, From Dispute Resolution to Dispute Avoidance: Some thoughts on collective management of treaty performance, in LIBER AMICORUM GUNTHER JAENICKE (VOLKMAR GÖTZ ET AL. EDS) 353-374 (1998), Thomas Gehring, International Environmental Regimes: Dynamic Sectoral Legal Systems 1 YIEL 353-374 (1990) and this author's, Enforcement and Countermeasures in the WTO: Rules are Rules - Towards a More Collective Approach, AMERICAN JOURNAL OF INTERNATIONAL LAW, Vol. 94, April 2000, pp. 621-633.
60 Commentary to the Final ILC Draft, pp. 320-1, para. (7). See also Crawford, Third Report, para. 92.
61 United States Diplomatic and Consular Staff in Tehran, I.C.J. Reports 1980, p. 3, at p. 43, para. 92. Quoted infra.
62 WOLFGANG BENEDEK, DIE RECHTSORDNUNG DES GATT AUS VÖLKERRECHTLICHER SICHT 468 (1990).
63 Appellate Body report on European Communities - Regime for the Importation, Sale and Distribution of Bananas, WT/DS27/AB/R, adopted on 25 September 1997, para. 136.
64 Ibid., para. 133.
65 Ibid., footnote 66.
66 As Philip Jessup put it in 1948: "[T]here must be basic recognition of the interest which the whole international society has in the observance of the law ... More broadly, the acceptance of the hypothesis of community interest should be considered to vest in all members of the international community a legal interest in respect for treaties ... Respect for the maxim pacta sunt servanda and the development of treaty law will be matters of concern to all states, and an infringement of the law will affect the interests of all" (PHILIP JESSUP, A MODERN LAW OF NATIONS 2, note 2, and 154 (1948)).
67 Nor is it the way the VCLT has been drawn up (i.e., it has been drawn up on the basis of bilateral state relations, not in light of community interests). See Shabtai Rosenne, Bilateralism and Community Interest in the Codified Law of Treaties, in TRANSNATIONAL LAW IN A CHANGING SOCIETY, ESSAYS IN HONOR OF PHILIP C. JESSUP (ED. W. FRIEDMANN, L. HENKIN AND O. LISSITZYN) 203-227 (1972).
68 See supra text at note 38.
69 Crawford, Third Report, para. 104.
70 Appellate Body report on EC - Bananas, paras. 136-137
71 Ibid., para. 138.
72 In support of the need for a locus standi doctrine in WTO dispute settlement see: Rodrigo Bustamante, The Need for a GATT Doctrine of Locus Standi: Why the United States Cannot Stand the European Community's Banana Import Regime, 6 MINN. J. GLOBAL TRADE 533-583 and William Davey, Has the WTO Dispute Settlement System Exceeded its Authority?, 2001 JIEL 95, 97-99.
73 Hence, even WTO transparency provisions or DSU obligations prohibiting legislation calling for the unilateral enforcement of WTO rights, outside of the DSU mechanism, are not integral in nature. Measures in violation of these obligations may well, on the books, apply to all other WTO members as well as all imports so that the first condition for standing is easily satisfied. But under the second condition, a member wanting to challenge a violation of, say, GATT Art. X on publication of certain trade measures or US Section 301 allegedly imposing the unilateral enforcement of US rights under the WTO in violation of the DSU, must show the existence of some trade flows or at least trade potential being effected by these measures. In practice, this second condition will be easily met by most WTO members, assuming that they all have some trade flows or at least some trade potential in their relationship with the member imposing the measure. Nonetheless, the fact that all WTO members may then have standing to challenge these measures does not flow from their integral nature, but from the fact that all WTO members are affected in their bilateral legal relationship with the member allegedly breaching WTO rules.
74 See supra note 36.
75 As the Commentary to the Final ILC Draft Articles put it (with reference to the case on United States Diplomatic and Consular Staff in Tehran, ICJ Reports 1980, p. 3, at p. 43, para. 92): "The identification of one particular State as injured by a breach of an obligation under the Vienna Convention on Diplomatic Relations does not exclude that all States parties may have an interest of a general character in compliance with international law and in the continuation of international institutions and arrangements which have been built up over the years" (Commentary, p. 298, para. (9).
76 Another reason not to allow complaints against a measure that does not cause nullification to the complainant in question is that otherwise a claim could succeed on purely legal grounds but once implementation does not follow, the complainant (having a legal interest only) would not be able to retaliate since retaliation must be "equivalent" to the nullification caused (DSU Art. 22.4). If there is no nullification, the complainant could then not suspend any concessions in retaliation.
77 See, for example, the Panel report on US - Taxes on Petroleum and Certain Imported Substances, adopted on 17 June 1987, BISD 34S/136, para. 5.2.2 referring to GATT Art. III as a provision "not only to protect current trade but also to create the predictability needed to plan future trade".
78 See GATT Art. XXIII. Note, in this respect, that the Appellate Body found also that for procedural objections in respect of a panel ruling to be upheld, the member invoking them must show prejudice (Appellate Body report on European Communities - Measures Affecting Livestock and Meat (Hormones), WT/DS48/AB/R, adopted on 13 February 1998, footnote 138).
79 Appellate Body report on India - Measures Affecting Imports of Woven Wool Shirts and Blouses, WT/DS33/AB/R, adopted on 23 May 1997, p. 13.
80 Appellate Body report on EC - Bananas, para. 251.
81 As long as such deviations are not prohibited by the WTO treaty itself (such as regional arrangements not meeting GATT Art. XXIV) and do not affect the rights of third parties. See supra section I.A.3. Art. 41.1 VCLT provides as follows:
"Two or more of the parties to a multilateral treaty may
conclude an agreement to modify the treaty as between themselves alone if:
the possibility of such modification is provided for by the treaty; or
(b) the modification in question is not prohibited by the treaty and:
does not affect the enjoyment by the other parties of their rights under the treaty or the performance of their obligations;
does not relate to a provision, derogation from which is incompatible with the effective execution of the object and purpose of the treaty as a whole" (emphasis added).
For a discussion of Art. 41 VCLT as it applies to the WTO treaty, see this author's, The Role of Public International Law in the WTO: How Far Can We Go?, AMERICAN JOURNAL OF INTERNATIONAL LAW, Vol. 95, July 2001, 535, at 547-50.
82 Arts. 30.3-5 VCLT provide as follows:
3. When all the parties to the earlier treaty are parties also to the later treaty but the earlier treaty is not terminated or suspended in operation under article 59, the earlier treaty applies only to the extent that its provisions are compatible with those of the later treaty.
4. When the parties to the later treaty do not include all the parties to the earlier one:
(a) as between States Parties to both treaties the same
rule applies as in paragraph 3;
(a) as between a State Party to both treaties and a State Party to only one of the treaties, the treaty to which both States are parties governs their mutual rights and obligations.
5. Paragraph 4 is without prejudice to article 41, or to any question of the termination or suspension of the operation of a treaty under article 60 or to any question of responsibility which may arise for a State from the conclusion or application of treaty, the provisions of which are incompatible with its obligations towards another State under another treaty" (emphasis added).