Marko Divac Öberg, The Legal Effects of Resolutions of the UN Security Council and General Assembly in the Jurisprudence of the ICJ, European Journal of International Law, Volume 16, Issue 5, November 2005, Pages 879–906, https://doi.org/10.1093/ejil/chi151
Navbar Search Filter Mobile Enter search term Search Navbar Search Filter Enter search term SearchThis article aims to extract from the jurisprudence of the International Court of Justice a basic theory of legal effects of unilateral instruments of international organizations in public international law. These effects can be divided into three categories. The first is substantive effects. These include binding, authorizing and (dis)empowering effects. The second category is causative effects, whereby determinations of fact or of law bring substantive effects into existence. The third category is modal effects – how and when the substantive effects come into existence (e.g. immediate or deferred, retroactive or non-retroactive, reversible or irreversible effect). Each of these categories of legal effects behaves differently according to whether the effects are intrinsic or extrinsic. Intrinsic effects are based on the special treaty powers of the United Nations Security Council and General Assembly. In this hypothesis, all three categories of effects exist to the full extent that the explicit and implicit powers of the adopting body allow for them. Extrinsic effects are directly based on general international law, in particular on the rules of formation of customary international law. Here, there are no causative effects. Substantive effects do not strictly speaking exist; only pre-substantive ones do. And modal effects are always immediate, non-retroactive and reversible.
Previous studies of the legal effects of resolutions of the United Nations Security Council (SC) and General Assembly (GA), as established in the judgments and opinions of the International Court of Justice (ICJ), have focused on binding effect, with only passing references to other substantive effects such as authorizing effect and (dis)empowering effect or to the modal effects that shape them and the factual and legal determinations that trigger them. 1 This article aims to correct that imbalance.
The effects differ according to the type of resolution. 2 The term ‘resolution’ as used in UN practice has a generic sense, including recommendations and decisions, both of which have a vague and variable meaning in the Charter. 3 The Court, on the other hand, reserves the expression ‘decision’ for binding resolutions and ‘recommendation’ for non-binding ones. 4 A resolution is ‘binding’ when it is capable of creating obligations on its addressee(s). 5 There is some disagreement over whether declarations, which in theory only interpret the Charter or assert the content of general international law, 6 constitute a sub-category of recommendations or a separate category. Our analysis will show that there is a point in treating these as a separate category. Note that a resolution, as a formal instrument, may combine different provisions that, substantively, respectively recommend, decide or declare. These three expressions will here be used in their substantive meaning, whereas ‘resolution’ will, depending on the context, either be a generic substantive term or designate the formal instrument.
Other factors relevant for the effects are the conventional 7 or customary 8 legal basis of the resolutions, their compatibility with the Charter (intra vires or ultra vires), 9 their addressees (one member, some members, all members, other UN organs . ), their subject matter (to which the Charter may attach different legal consequences), their terminology (shall as opposed to should, recommend as opposed to demand, etc.), and, for the possible effects on international customary law, the ways they are adopted, who and how many vote for and against them, and perhaps even why they do so. 10 But the title of the resolutions (declaration, code, charter. ) is irrelevant, as is the express or implied nature of the powers upon which their adoption is based. 11
The most fundamental factor is whether the effects are intrinsic or extrinsic. Intrinsic effects stem directly and immediately from the adoption of the resolution, on the basis of powers supplied by a treaty or the customary law internal to it (usually the UN Charter, but possibly another treaty making use of the existing UN institutional structure 12 ). Extrinsic effects spring from the resolution but are, due to the adopting body’s lack of the necessary powers, directly based on international customary law. 13 The difference between the two hypotheses is the absence or presence, between the resolution and general international law, of an intermediate legal basis providing the adopting body with the relevant special powers.
There are three basic types of legal effects. 14 A legal rule, when triggered by a determination that the conditions for its application are fulfilled, states the obligations, rights and powers that result. 15 A resolution may therefore have the legal effect of (i) creating obligations, rights and/or powers (which we shall call ‘substantive effects’) 16 , and/or (ii) making determinations 17 of facts (e.g. that an alleged fact is true) or legal situations (e.g. that an obligation was violated), which trigger the substantive effects (‘causative effect’). To this should be added (iii) how and when the substantive effects operate (‘modal effects’). Each of these categories has a dual nature according to whether the effects are intrinsic or extrinsic. By showing this, the present article aims to contribute to the basic theory of the legal effects of unilateral instruments in public international law. 18
Several issues are closely related to the present topic, yet fall outside of it. Sometimes there is only an illusion of legal effects. This is the case when a resolution simply restates an obligation, a right or a power that already exists. Declarations in principle only interpret or restate the law, in which case they have no legal effect. Likewise, a resolution which merely interprets the Charter does not, in theory, have any legal effect of its own. To the extent that it details and substantially adds to the Charter, any ensuing legal effect does not come from the resolution of a given organ, but from the fact that it may be considered generally acceptable by UN Members. 19 Here we find legal effects, but they do not originate in the resolution. Legal effects deriving from someone’s (anticipatory or subsequent) acceptance of a resolution, or the particular way in which it was adopted, or obligations protracted on its basis, 20 do not stem from the resolution itself.
Section 1 will examine the intrinsic effects of decisions and recommendations on the UN legal order. The possible extrinsic effects of declarations on general international law are then considered in Section 2.
The ICJ has not recognized any intrinsic legal effects based on customary norms internal to the UN legal order or operating on general international law. 21 Hence, this section is limited to effects based on treaty law and operating on the UN legal order. Most of the Court’s discussions of the legal effects of GA and SC resolutions have concerned the existence, force and scope of binding effect. But it has also dealt with authorizing effect, which is not necessarily the mirror image of binding effect, and (dis)empowering effect. Finally, the Court has begun to outline its approach to the causative and modal effects respectively triggering and shaping the substantive ones.
Discussions of binding effect abound in ICJ jurisprudence and legal literature. Consequently, this section will only provide a concise overview. Only decisions have binding effects; recommendations do not. 22 Crudely put, the decisional powers of the GA are restricted to ‘organizational’ matters internal to the UN legal order (including semi-external matters such as the budget, or admission, suspension and expulsion of members), while the SC also possesses decisional powers in the ‘operational’ realm of international peace and security. 23
The binding effect of GA decisions is limited, ratione materiae, to organizational matters, but may cover, ratione personae, the entire UN sphere.
Although GA resolutions are recommendatory as a rule, 24 especially regarding external relations with Member States, 25 the Court has recognized the binding legal effect of GA decisions pertaining to the admission of new Member States, 26 voting procedure, 27 or apportionment of the budget, 28 and in general has confirmed that the Court possesses certain powers of decision. 29 The Court has never clarified whether the GA has any decisional powers in mandate/trusteeship matters. 30 Resolutions of the GA have no binding effect in the operational realm of international peace and security. Neither the GA’s budgetary powers in this area, nor its enforcement powers to suspend or expel UN Members, fall outside of the organizational sphere. 31
Ratione personae, GA decisions obviously bind their (valid) addressees. They may also bind the UN at large, and consequently all Member States, e.g. through their regular contributions to the budget. 32 This generalized effect includes those that voted against the decision, such as the trustee state in questions pertaining to its trusteeship. 33 So the binding scope of GA decisions covers the entire internal UN sphere.
The ICJ has not definitively decided whether SC decisions possess an overriding binding effect, but it has specified that the binding effect includes, ratione materiae, operational matters and covers, ratione personae, all Member States.
Unlike the recommendations of the SC, 34 its decisions have binding force, 35 but the Court has made only a provisional finding that SC decisions have an overriding normative power capable of pre-empting obligations flowing from traditional sources of international law. 36 Recognizing such overriding binding force would give a secondary source of UN law (decisions) a greater normative value than many primary sources of international law (treaties) – thereby giving the SC a potentially very disruptive power – and would ultimately place great faith in the SC truly acting on behalf of all Member States. 37
Ratione materiae, the binding effect of SC resolutions belongs to the realm of international peace and security 38 and includes enforcement under Chapter VII of the UN Charter, 39 but is not limited to that. 40 Since just about any significant international event or situation can be characterized as a threat to peace and security, 41 the scope of the SC’s binding powers, if combined with an overriding binding force, would make the SC a dauntingly powerful organ. Whether a specific SC resolution is binding is determined by the language used in it, the discussions leading to it, the Charter provisions invoked, etc., 42 all with the purpose of establishing the intent of the SC. 43 The precise content of the binding effect is left to the SC itself, 44 but the Court has found certain ‘implicit’ legal effects and, inversely, put some limits on the effects 45 when these conflict with the principles and purposes in Chapter I of the UN Charter. 46 This limitation is too vague to have much practical value in the absence of any organ competent to review the validity of SC resolutions. 47
Ratione personae, an SC decision may bind all UN Member States, including ‘those members of the Security Council which voted against it and those Members of the United Nations who are not members of the Council’. 48 As for non-Member States, the most coherent interpretation of a difficult passage in the Namibia opinion rejects any direct binding effect. 49 This interpretation respects the basic principle that treaties only bind parties, and avoids the difficult question of whether the UN Charter is subject to special rules within the law of treaties. It also leads to the same practical outcome since just about every state is now a member of the UN.
Some resolutions create rights, implying reciprocal obligations to fulfil those rights or at least not to interfere with them. 50 But are authorizations then necessarily included in decisions, or may they also be included in recommendations?
The Court has on several occasions endorsed the idea of a UN decision having an authorizing effect.
The 1962 Certain Expenses opinion noted: ‘If the Security Council, for example, adopts a resolution purportedly for the maintenance of international peace and security and if, in accordance with a mandate or authorization in such resolution, the Secretary-General incurs financial obligations, these amounts must be presumed to constitute “expenses of the Organization”’. 51 The reference to ‘the maintenance of international peace and security’ suggests that the Court found the given resolution to be a decision. The ICJ also quotes the SC resolution establishing the United Nations Operations in the Congo (ONUC), according to which the Security Council ‘Decides to authorize the Secretary-General. ’. 52 Clearly, an SC decision here had an authorizing effect.
The Court has repeatedly found that it has a discretionary ‘power’ to accept or refuse UN requests for ICJ advisory opinions. 53 Since the Court cannot issue opinions sua sponte, the legal effect of the GA resolution is that of an authorization. In line with the rather categorical ‘request’ for an advisory opinion found both in the UN Charter (Article 96) and the ICJ Statute (Articles 65 and 66), ICJ practice shows that these resolutions are decisions. 54 The Court has found a number of qualifications to its discretionary right, none of which alter this analysis. 55 The Court has even given a broad scope to the authorizing effect, finding that it must not be unduly restrained by the wording of the request or the circumstances surrounding its adoption. The Court ‘must have full liberty to consider all relevant data available to it in forming an opinion on a question posed to it for an advisory opinion’. 56
It is natural that decisions may have authorizing effects, since an authorizing effect upon one person corresponds to a ‘correlative’ binding effect upon one or several other persons, the content of which is to accommodate that right.
More puzzlingly, the ICJ seems to have admitted that recommendations may have at least a certain form of authorizing effect.
One purported example of this is better explained in a different manner. Article 4(2) of the UN Charter provides that admission to the UN ‘will be effected by a decision of the General Assembly upon the recommendation of the Security Council’. In its 1950 Admission to the United Nations opinion, the Court had to decide whether a General Assembly decision could admit a state when the Security Council had not transmitted a recommendation to it. It found that ‘[b]oth these acts are indispensable to form the judgment of the Organization’, since ‘the recommendation of the Security Council is the condition precedent to the decision of the Assembly by which the admission is effected’. 57 The Court does not pronounce itself expressly upon the legal effect of the SC recommendation. If it were an authorization, 58 then it would incur reciprocal obligations not to interfere with the GA’s discretionary choice to invite or not a new Member State. But a more likely interpretation is that the SC recommendation simply makes a necessary determination that the applicant fulfils all the requirements of article 4(1) of the UN Charter. 59
The 1962 Certain Expenses opinion, on the other hand, does appear to confirm that recommendations may have authorizing effects. The Court found the GA resolutions establishing the United Nations Emergency Force in the Middle East (UNEF) to be ‘recommendations’ taken under Article 11 or 14 of the UN Charter. 60 But a resolution cannot establish a physical reality like UNEF by itself, it can only formulate a new legal situation. 61 Then the legal effect of the above-mentioned recommendations must be to authorize the UN Secretary-General to establish UNEF. 62 The right of the Secretary-General to establish the force implies reciprocal obligations not to interfere with that right. 63 But the ICJ has otherwise consistently reserved the expression ‘recommendations’ for non-binding resolutions. Did the Court really intend to stray from this in the Certain Expenses case?
There may be a way to explain this apparent contradiction. Neither of the cases involving a possible authorizing effect in recommendations implied a reciprocal affirmative duty on any specific addressee. Rather, they implied general reciprocal obligations not to interfere with the given right. Is this then a true ‘right’? W. N. Hohfeld distinguishes between rights and duties on the one hand, and privileges and ‘no-rights’ on the other. 64 Applying this distinction to the Certain Expenses example, the UN Secretary-General has a ‘privilege’ (as opposed to a ‘right’) to create UNEF, which is correlative of a series of ‘no-rights’ (rather than duties) to prevent the Secretary-General from doing so. 65 So perhaps recommendations can only create privileges and not actual rights. More radically, H. Kelsen argues that when an obligation is not owed to a specific determinable person but to a general community, then there may be no corresponding right. 66 If the argument is turned around, a right not implying a reciprocal obligation on any specific person may not have any corresponding obligation at all. Under this analysis, the GA recommendations authorizing the creation of UNEF did not have any kind of binding effect and so did not prevent Member States from interfering with the Secretary-General’s right. In both theoretical constructions, the crucial point for our purposes is that the ‘right’ has no active reciprocal duty, and exists towards a general community rather than a specific person. A final possibility is that the Court reached an unsound conclusion.
The Court has only on rare occasions dealt with the empowering effects of GA and SC resolutions. Its discussions nonetheless show that empowering and disempowering effects are opposites that may cancel each other out, although sometimes the (dis)empowering organ may not be able to take back what it has given.
The 1954 UNAT opinion discussed the (dis)empowering effects of GA decisions:
There can be no doubt that the Administrative Tribunal is subordinate in the sense that the General Assembly can abolish the Tribunal by repealing the Statute, that it can amend the Statute and provide for review of the future decisions of the Tribunal and that it can amend the Staff Regulations and make new ones. There is no lack of power to deal effectively with any problem that may arise. But the contention that the General Assembly is inherently incapable of creating a tribunal competent to make decisions binding on itself cannot be accepted. 67
So the GA decision establishing the United Nations Administrative Tribunal (UNAT) empowered it to create binding effects on the GA itself. 68 But the Court notes that UNAT remains subject to the (dis)empowering effects of GA decisions, for instance by amending or repealing the statute of UNAT. Empowering and disempowering effects are therefore opposites, one of them annulling the other in the same person. Binding and authorizing effects, on the other hand, are not only opposites but also correlatives, meaning that a binding effect upon one person corresponds to an authorizing effect upon someone else. It is somewhat artificial to think of (dis)empowering effects having correlative effects subjecting a person to – or subtracting that person from – the power of another, since a power only has latent effects upon its subjects (realized only if the power is exercised). But are there any limits on the ability of empowering and disempowering effects to cancel each other out?
The passage cited above in the UNAT case, according to which ‘the General Assembly can abolish the Tribunal by repealing the Statute’, is in a state of tension with the Court’s finding in the same case
that the power to establish a tribunal, to do justice as between the Organization and the staff members, was essential to ensure the efficient working of the Secretariat, and to give effect to the paramount consideration of securing the highest standards of efficiency, competence and integrity. Capacity to do this arises by necessary intendment out of the Charter. 69
This passage suggests that the Charter intended the creation of UNAT. A power that is ‘essential’ to the workings of the UN and a ‘necessary’ implication of the Charter seems like one that should be exercised. If so, the GA ought not to adopt a resolution with a disempowering effect so extensive as to take back all powers previously conferred on the UNAT. Yet, the Court clearly said that it can do it. The ICJ seems to have grappled with the problematic implications of the Charter not providing for UNAT. Anyhow, all possible limits on the GA’s future disempowerment of the UNAT would flow from the Charter itself and not from the GA’s initial empowerment of the UNAT.
On the other hand, the GA can disempower itself and the UN in general to the point of taking away the very power that allows it to do so. In the 1963 Northern Cameroons case, the Court found that ‘the General Assembly is no longer competent pursuant to the termination of the Trusteeship as a result of resolution 1608 (XV)’, which caused all competent UN organs to lose their trusteeship powers in relation to the given territory. 70 The GA could not regain its powers by passing a resolution reversing the previous one. 71 This reflects the crucial difference between binding/authorizing effects and (dis)empowering ones – the latter confer or eliminate the very capacity to create legal effects.
Empowering and disempowering effects, like their binding and authorizing cousins, may cancel each other out. But empowerment, being the capacity to create obligations, rights and powers, means that there are also inherent differences. If the UNAT could bind the hierarchically superior GA, surely it could not empower it. When the GA disempowered the UN entirely with regard to the Northern Cameroons trusteeship, it also lost the power to re-empower itself. The provisions found to have (dis)empowering effects were decisions, never recommendations.
Causative effect triggers, through determinations, dormant substantive effects. A factual determination is a finding that, in the eyes of the determining body, a certain fact did or did not happen. This may or may not be legally relevant, but only if it is legally relevant is there any causative effect. It is, for example, a (legally relevant) factual determination that one state has invaded the territory of another. A legal determination is sometimes explicitly or implicitly based on a factual determination which the determining body characterizes in legal terms (for instance, if the SC finds that a state has committed an act of aggression under Article 39 of the UN Charter), sometimes it is merely an identification of a specific legal situation (for instance, determining whether a particular state is or is not a trustee). Determinations have legal effects by blocking or causing the applicability of certain rules, thereby triggering substantive effects. 72 For instance, the determination by the SC that there is a ‘threat to the peace, breach of the peace, or act of aggression’ marks the end of the (exclusive) applicability of Article 33 of the UN Charter and the obligations that it confers. 73
Two problems have surfaced in the ICJ’s jurisprudence pertaining to when GA and SC determinations are binding. 74 First, do determinations take on the binding or non-binding effect of the resolutions that they are included in? Second, do these determinations bind the ICJ?
The ICJ has addressed the question of the binding effect of determinations both in recommendations and decisions, apparently giving them the force of the resolutions containing them.
Determinations made in recommendations appear to be non-binding. In a previously quoted passage of the 1950 Admission to the United Nations opinion, the Court found that, for an admission to the UN, both a General Assembly decision and a Security Council recommendation ‘are indispensable to form the judgment of the Organization’, since ‘the recommendation of the Security Council is the condition precedent to the decision of the Assembly by which the admission is effected’. 75 The most natural interpretation is that the SC recommendation makes a determination that the applicant fulfils all the requirements of Article 4(1) of the UN Charter, 76 upon which the GA can then base its decision. The determination is not binding, since the GA can decide to admit the state or not. This matches the non-binding character of a recommendation. The causative effect is optional, within the discretion of the body that applies the norm.
Inversely, determinations made in decisions appear to be binding. In the Namibia case, the Court found that ‘[a] binding determination made by a competent organ of the United Nations to the effect that a situation is illegal cannot remain without consequence . there is an obligation, especially upon Members of the United Nations, to bring that situation to an end’. 77 Although the Court offers no explanation, the legal determination was presumably binding because it was included in a decision contained in SC Resolution 276. The causative effect is binding upon the states that apply the norm.
All in all, it seems that the binding or non-binding nature of a resolution (decision or recommendation) also covers determinations made therein; a determination made in a recommendation is not binding, whereas a determination made in a decision is. 78 This parallelism prevents, in case of overlapping competences, an organ with recommendatory competence from pre-empting the determinations of an organ with decisional competence. But is there a specific exception, ratione personae, to this rule?
It is in the nature of the ICJ, as a court, to make determinations. If a determination on a specific issue has already been made by one of the other principal UN organs, should the ICJ be in a position to challenge it? The Court has approached this question, but never squarely decided it.
Looking to establish the relevant facts in the 1986 Nicaragua case, the Court found that ‘in its quest for the truth, it may also take note of . the resolutions adopted or discussed by [international] organizations, in so far as factually relevant’. 79 This could be read to imply that factual determinations in GA and SC resolutions do not impose themselves upon the Court. However, the cited passage deals with the probative, not the binding, value of such determinations, without distinguishing between decisions and recommendations, or even between UN resolutions and resolutions of other international organizations. One cannot conclude from this that the Court is free to reconsider determinations made in UN resolutions.
Indeed, the East Timor case could support the opposite proposition. Portugal argued that the Court must take as ‘givens’ the GA and SC resolutions making the legal determination that East Timor was a non-self-governing territory and that Portugal was the administering Power thereof. The Court found that the resolutions did not intend to impose any obligations, noted some conflicting state practice and concluded that the determinations in the resolutions therefore could not be regarded as ‘givens’ sufficient for deciding the dispute. 80 This leaves open the possible interpretation that, a contrario, legal determinations in binding decisions could be ‘givens’ imposing themselves upon the Court. But such reasoning is unreliable, since the Court may never find that any such determinations constitute ‘givens’.
All in all, it is not clear from the Court’s jurisprudence whether it is free to reconsider factual and legal determinations made in GA and SC resolutions for the purposes of settling a dispute. 81 These determinations may have been made on the basis of partial information, where not all interested parties were heard, and/or too urgently for the facts to be objectively established. 82 Although it may be tempting to rely on what other principal organs of the UN have found, the ICJ, with its judicial nature and careful approach to establishing facts, should not be bound by such determinations.
Modal effects establish how and when substantive effects operate. 83 The ICJ’s jurisprudence is still rather undeveloped on this point. 84 It has made a few findings about the temporal aspects of substantive effects, as well as their reversibility.
The ICJ has implicitly approved of a GA decision deferring its own effect, but has denied any retroactive effect to both GA and SC decisions.
The ICJ dealt with immediate and deferred legal effects in the complex Mortished case. Mr. Mortished retired as a UN staff member on 30 April 1980. GA decision 33/119 of 19 December 1978 had recently changed the rules on the payment of repatriation grants to staff members leaving UN service, subjecting it to evidence of relocation. The Secretary-General amended Staff Rule 109.5 accordingly, paragraph (f) of which provided that staff members already in service before 1 July 1979 would retain their accrued right to a repatriation grant without providing evidence of relocation. This applied to Mr. Mortished.
But on 17 December 1979, the GA adopted decision 34/165, effective as of 1 January 1980, according to which all staff members would be subject to the evidence requirement. No longer exempt, Mr. Mortished’s seised the UNAT, which ruled that he held an acquired right by virtue of Staff Rule 109.5(f), even though it was no longer in force at the date of his retirement. Acquired rights were protected by Staff Regulations and Rules, entitling Mr. Mortished to compensation. The Committee on Applications for Review of Administrative Tribunal Judgments requested an advisory opinion from the ICJ on whether the UNAT had erred in not giving immediate effect to GA decision 34/165’s requirement of evidence of relocation.
The ICJ found that UNAT’s ‘decision was not that resolution 34/165 could not be given immediate effect but, on the contrary, that the Applicant had sustained injury precisely by reason of its having been given immediate effect by the Secretary-General in the new version of the Staff Rules which omitted Rule 109.5(f).’ 85 The ‘immediate effect’ given by the Secretary-General does not arise from the resolution itself, but from the Staff Rules. Indeed, GA decision 34/165, adopted on 17 December 1979, did not have an immediate effect, since it deferred its own substantive effect until 1 January 1980. 86 Deferred effect is no doubt the exception and immediate effect the rule, since it is hard to imagine any other reasonable default starting point for a substantive effect.
The ICJ has clearly denied retroactive effect to both GA and SC resolutions. In the preliminary objections phase of the Lockerbie case, the Court had to decide whether the case was inadmissible because certain SC resolutions had disposed of the dispute. The Court denied the objection, finding that ‘[t]he date, 3 March 1992, on which Libya filed its Application, is in fact the only relevant date for determining the admissibility of the Application. Security Council resolutions 748 (1992) and 883 (1993) cannot be taken into consideration in this regard since they were adopted at a later date’. 87 Even if the Court had found another date to be relevant, this argument denies any retroactive effect to the given resolutions.
The Court similarly addressed a GA decision in the 2003 judgment on the application for revision of the judgment on preliminary objections in the Genocide case. The Federal Republic of Yugoslavia (FRY) argued that its recent admission to the UN showed that at the time of the initial judgment it had not been a party to the ICJ Statute, had not been bound by the Genocide Convention and hence the ICJ had no personal jurisdiction over it. The Court found that ‘General Assembly resolution 55/12 of 1 November 2000 [admitting the FRY to the UN] cannot have changed retroactively the sui generis position which the FRY found itself in vis-à-vis the United Nations. ’. 88 Non-retroactive effect is no doubt the rule, as it is in all legal systems for the sake of legal security, but one may imagine cases in which a resolution could provide for its own retroactivity. 89
The ICJ has recognized the irreversibility of one kind of UN resolution, but also the reversibility of others.
In the 1963 Northern Cameroons case, the ICJ found that GA Resolution 1608 (XV) – terminating a trusteeship agreement – had a ‘definitive legal effect’ that a Court judgment could not change. 90 This ‘definitive legal effect’ must be understood as an exception rather than a rule. First of all, the termination of a trusteeship agreement by nature has a definitive effect that even the GA itself could not reverse, since Resolution 1608 (XV) disempowered the GA in the matter. 91 Moreover, the Nauru judgment reaffirmed the Northern Cameroons finding, also in the context of the termination of a trusteeship agreement, noting that ‘[s]uch a resolution had “definitive legal effect”’. 92 Other resolutions might not.
Indeed, under normal circumstances, the legal effects of a resolution must be understood to end where the adopting body reverses it by a new incompatible resolution. In the previously discussed Mortished case, GA Resolution 34/165 effectively reversed the prior decision in Resolution 33/119, which exempted some UN staff members from having to provide evidence of relocation in order to obtain their repatriation grant. Such reversible legal effect must be the rule, or else two incompatible substantive legal effects would coexist.
Under special circumstances, the legal effects are ‘definitive’, meaning that no change or reversal is possible. Under normal circumstances, another resolution by the same body may no doubt reverse them, partially or entirely.
These findings of the Court, incomplete as they are, reflect some basic truths about the modal effects attached to obligations, rights and powers created by GA and SC resolutions. As a rule, these are immediate, non-retroactive and reversible. In many cases the resolution itself could modulate the first two, by deciding on its own retroactivity, deferring its effect or, if the effect is not instantaneous, programming its own modification or termination (transient legal effect). 93 A resolution cannot, on the other hand, make itself irreversible since such a stipulation would have no greater legal force than the future stipulation that would reverse it. Irreversibility applies if the resolution itself disempowers the adopting body in the matter or if a non-Charter instrument empowers it to create a specific effect but not to take it back.
The jurisprudence of the ICJ is particularly rich on the binding effect of GA and SC decisions. No definitive differentiation has been made as to the force of the binding effect in each case. Determinations of facts and legal situations made in decisions are also binding. The scope of the binding effect is remarkably broad for both organs, though each within its specific sphere of competence. Whereas the GA’s Charter-based powers of decision are organizational, only the SC has operational powers of decision based on its responsibility for the maintenance of international peace and security. But one may doubt whether the SC’s sphere of competence constitutes any real limitation.
Somewhat less rich are the Court’s discussions of authorizing, empowering or disempowering effects. Decisions may have all of these. Less explicitly, the Court has recognized that recommendations may also have at least some kind of authorizing effect. In these cases, the Court dealt with authorizations that created generalized reciprocal obligations not to interfere with the exercise of the given right, rather than a specific reciprocal duty of action. So maybe recommendations cannot possess the full authorizing effect of creating a right that implies an affirmative duty on some specific addressee(s). In any event, it seems that recommendations do not contain binding determinations or have (dis)empowering effects.
Declarations have none of these effects. Declarations have not been discussed yet, primarily because they were not foreseen by the Charter and have no effect based on it. They either interpret the Charter or reflect the state of international law, in which case they have no effect of their own; or they add provisions of lex ferenda, in which case they have no Charter authority to influence general international law. But international customary law may provide an extrinsic basis for their effects. 94
In the previous section, the basis of the legal effects of the resolutions was conventional and they acted upon the internal UN legal order. In this section, the legal effects operate on general international law 95 and are based on customary law. They are extrinsic because, although it is declarations that have legal effects, these are directly based on international customary law. Between the two there is no intermediary instrument providing the adopting body with intrinsic powers.
The ICJ has dealt with three main issues concerning the effects on general (customary) international law: whether there are any such effects; on which of the two constitutive elements of customary law – opinio juris and practice – any such effects operate; and on the basis of which criteria they do so.
If a UN resolution merely interprets pre-existing substantive international norms, it may be helpful for understanding and applying them. If it restates existing international norms, it may have an evidentiary value for establishing these. But in neither case does the resolution have any impact on the state of the law. Granted, in practice it can be hard to draw the line between what, on the one hand, is merely interpretative or declaratory and what, on the other hand, is truly creative. The Court has often been vague in separating the impact of UN resolutions on customary law from their interpretative or evidentiary value. After having been entirely unclear in the 1970s, the Court gained in clarity in the 1980s and then retreated again in the 1990s.
In the 1970s, the Court identified GA declarations as a ‘further important stage’ in the development of international law, 96 or inferred the ‘existing rules of international law’ from them, 97 but made no mention of how or why this could be done. 98
The 1986 Nicaragua judgment achieved greater clarity. For instance, it found that the description of acts constituting armed attacks annexed to GA Resolution 3314 (XXIX) ‘may be taken to reflect customary international law’. 99 The word ‘reflect’ indicates that GA resolutions are here used as evidence of customary law 100 and are therefore not given any legal effect.
But the Nicaragua opinion also took a different approach, this time confirming that UN resolutions may have an impact on customary law. Searching for an opinio juris concerning the rule of abstention from the threat or use of force against the territorial integrity or political independence of other states, the Court found that:
This opinio juris may, though with all due caution, be deduced from, inter alia, the attitude of the Parties and the attitude of States towards certain General Assembly resolutions, and particularly resolution 2625 (XXV) entitled ‘Declaration on Principles of International Law concerning Friendly Relations and Co-operation among States in accordance with the Charter of the United Nations’. The effect of consent to the text of such resolutions cannot be understood as merely that of a ‘reiteration or elucidation’ of the treaty commitment undertaken in the Charter. On the contrary, it may be understood as an acceptance of the validity of the rule or set of rules declared by the resolution by themselves. 101
Here, the role of UN resolutions is indeed to participate in the creation of customary law and is neither confined to restatement or interpretation (‘reiteration or elucidation’) of the Charter, nor to mere evidence of the content of international customary law (since the ‘effect of consent to the text of such resolutions’ is ‘an acceptance of the validity of the rule or set of rules declared by the resolution’). The latter point is also supported by the fact that the Court does not discriminate between provisions based on lex lata and those based on lex ferenda.
However, the reference to the ‘attitude of States’ could mean that the effect does not come from the resolutions themselves but rather from the way states receive them. But the Court later adds, on the topic of Resolution 2625 (XXV): ‘As already observed, the adoption by States of this text affords an indication of their opinio juris as to customary international law on the question’. 102 This suggests that the previously mentioned ‘attitude of States’ towards the resolutions boils down to their adoption of them. 103 So the resolutions themselves do have an effect. 104 However, the Court arguably weakens its statement, firstly, by adding that the resolution offers a mere ‘indication’ of this opinio juris and, secondly, by listing several other vehicles for opinio juris, for intance statements by state representatives. 105
Unfortunately, the 1996 Nuclear Weapons opinion, while dissipating these lingering doubts, also created new ones. Searching for a customary rule prohibiting recourse to nuclear weapons, ‘[t]he Court notes that General Assembly resolutions, even if they are not binding, may sometimes have normative value. They can, in certain circumstances, provide evidence important for establishing the existence of a rule or the emergence of an opinio juris’. 106 Here, the Nicaragua reference to the attitude of states towards certain General Assembly resolutions is gone, as is the blurring statement that this attitude is a source ‘inter alia’ of opinio juris. But the citation also seems to conflate the two Nicaragua approaches (evidence of customary law/effect on opinio juris) into one by using the resolutions to provide ‘evidence’ of both, thereby eliminating any legal effect. On the other hand, the Court finds that GA resolutions may have ‘normative value’. The renewed lack of clarity is perhaps linked to persisting doubts about the relevance of UN resolutions for customary law. 107 The remainder of this article will assume the Nuclear Weapons opinion to be relevant for the question of legal effect. 108
Scholars have disagreed widely on whether UN resolutions may be constitutive of state practice or opinio juris. 109 Although one may argue that what states do is more important that what they say, 110 it can be difficult to separate words and actions in state practice. 111 State practice is sometimes what the state says (for instance, threatening to use a nuclear weapon, recognizing a state), sometimes what it does (for instance, using a nuclear weapon, abstaining from a vote). The Court has so far settled this issue in favour of opinio juris, at the exclusion of state practice.
As we have seen, both the 1986 Nicaragua decision and the 1996 Nuclear Weapons opinion expressly link GA resolutions to opinio juris. According to the former, ‘opinio juris may, though with all due caution, be deduced from, inter alia, the attitude of the parties and the attitude of states towards certain General Assembly resolutions’. 112 The latter found ‘that General Assembly resolutions, even if they are not binding, may sometimes have normative value. They can, in certain circumstances, provide evidence important for establishing the existence of a rule or the emergence of an opinio juris’. 113 So GA resolutions may inform opinio juris, but may they also constitute state practice?
The Nicaragua judgment, being extremely brief on the issue of state practice, does not clearly answer whether GA resolutions may also constitute state practice. 114 According to a number of authors, the answer is yes. 115 This view finds no explicit support in the decision, 116 and is rejected by other authors. 117 There are indeed other possible explanations why the Court is so brief on the topic of state practice. The Court may have thought it unnecessary and/or too difficult to examine the practice behind rules of abstention of such fundamental importance as non-aggression and non-intervention. 118 It does note some material state practice conflicting with the principle of non-intervention, which was not supported by an opinio juris capable of creating a new rule or exception. 119 Perhaps this shift reflected the practical necessity of examining action rather than abstention. The Court may also have been prevented – or thought itself to be precluded – from further examining state practice, due to most states being non-parties to the proceedings and due to the absence of the USA and hence of any material it could have submitted. 120 Finally, one may argue that the Court could hardly intend to use a single source for both constitutive elements of custom. This would give the resolutions instant quasi-legislative effect, 121 which finds no explicit support in the Charter and gives the GA and/or the SC an exorbitant power that was never intended. 122 It would also run the risk of ignoring how states behave in reality. 123
The Nuclear Weapons opinion interestingly considers that the ‘emergence, as lex lata, of a customary rule specifically prohibiting the use of nuclear weapons as such is hampered by the continuing tensions between the nascent opinio juris on the one hand, and the still strong adherence to the practice of deterrence on the other’. 124 Opinio juris and state practice are clearly separated. The debates in the Nicaragua and Nuclear Weapons cases concerned conceptually similar rules of abstention, prohibiting respectively (i) the threat or use of force against the territorial integrity or political independence of other states, and (ii) the threat or use of nuclear weapons. It is therefore submitted that the Nuclear Weapons case resolves the Nicaragua ambiguity as to whether UN resolutions could constitute not only opinio juris but also state practice. The answer is no, since state practice in the former case is confined to the policy of deterrence. But in the future the Court may still look to UN resolutions for State practice concerning rules that involve no material action (for example, state recognition).
The Nicaragua opinion was very vague on how opinio juris may be deduced from one or more General Assembly resolutions, 125 but the Nuclear Weapons opinion substantiated this: ‘it is necessary to look at its content and the conditions of its adoption; it is also necessary to see whether an opinio juris exists as to its normative character. Or a series of resolutions may show the gradual evolution of the opinio juris required for the establishment of a new rule’. 126
This quote identifies four criteria – conditions of the adoption, content, reality of the opinio juris and, perhaps, repetition, to which one more implicit criterion may be added, that the resolution be a declaration.
The expression ‘conditions of its adoption’ presumably refers to issues such as the resolution’s number of affirmative votes, abstentions or contrary votes, the representative nature of the states voting for or against it, and whether its mode of adoption permits a fair expression of each state’s point of view.
The Nuclear Weapons opinion finds that the relevant resolutions ‘fall short of establishing the existence of an opinio juris’, because ‘several of the resolutions under consideration in the present case have been adopted with substantial numbers of negative votes and abstentions’. 127 There might still be, according to the Court, an emerging ‘customary rule specifically prohibiting the use of nuclear weapons’, based on factors such as ‘the adoption each year by the General Assembly, by a large majority, of resolutions recalling the content of resolution 1653 (XVI)’. 128 Large majorities are thus crucial. This satisfies the requirement of a generalized opinio juris.
Nothing explicit is said about the importance of the representative quality of the supporting states, but the Court does give weight to the states engaged in a policy of nuclear deterrence. 129 It is reasonable that those states which are actually engaged in a certain activity have a strong say in how that activity is regulated.
The Court did not check whether the mode of adoption of the resolution allowed a fair expression of each state’s point of view. Yet this is not always the case, as illustrated by the GA’s practice of adopting resolutions by ‘false consensus’, without putting them to a vote that would reveal abstentions and contrary votes. 130
The Nicaragua judgment did not provide any explicit insights into the importance of the content of the resolutions, but one may draw some implicit conclusions. The Court seems to have given little weight to the context of the provisions that it finds to express opinio juris. Yet, Resolution 2625 (XXV) was based upon the Charter and not a statement about international customary law. 131 On the other hand, the provisions on which the Court relied used legal and mandatory language.
The Nuclear Weapons opinion summarily refers to the criterion of the ‘content of the resolution’. This presumably means whether the resolution has normative content and, if so, whether it is couched in legally binding terms. The Court noted normative language in the General Assembly resolutions put before it, some of which was clearly legal (‘violation of the Charter’). 132 Moreover, it analysed the legalistic content of Resolution 1653 (XVI) to conclude that there was no opinio juris for a specific customary rule prohibiting the use of nuclear weapons. 133
Judging from the two cases, the normative and legal language of a resolution’s relevant provisions is more important than their context. This is hard to justify – words taken out of context can be made to say many things.
The Nicaragua decision mostly just checks that the parties to the dispute had supported the given resolutions, without considering whether this is sufficient to constitute an opinio juris. But the Court did take into account a US statement that the declaration contained in GA resolution 2131 (XX) was ‘only a statement of political intention and not a formulation of law’, though the Court ultimately disregarded this because the similar principles in GA Resolution 2625 (XXV) had met with no such US statement. 134
The Nuclear Weapons opinion, on the other hand, states that, in order to find opinio juris in a GA Resolution, it is ‘necessary to see whether an opinio juris exists as to [the resolution’s] normative character’. 135 To transcend the tautology, this must either concern whether the adopting states believed that the normative content of the resolution was of a legal nature (psychological aspect) 136 or, perhaps, whether any normative content of the resolution is couched in legally binding terms (material aspect). 137 The Nuclear Weapons opinion provides no express clarification on this issue, but since the Court’s treatment of the content criterion is close to the latter alternative, the former seems more likely here. 138
In any event, the Court has recognized the importance of checking for the reality of the purported opinio juris. The GA has the attractive quality of being very broadly representative of the existing states, as well as constituting a centralized, highly convenient means of simultaneously identifying the points of view of all present Member States on a specific topic. 139 However, the GA is also a political organ, which does not make it an ideal forum for establishing the law. 140 States may indeed have reasons other than legal ones for voting the way they do, 141 such as moral, political, or pragmatic (for instance, as part of a bargain deal). Moreover, a state may vote against a resolution because it finds that it goes too far, or not far enough. Besides, it is hardly fair to bind a state to a favourable vote, when states ‘act within certain rules and mechanisms that normally affect the legal meaning of their votes’ and when resolutions are imputed not to individual members but to the adopting body and organization. 142 Finally, the state representatives who vote in the Assembly usually do not have the power to legally commit their states. The reality of the opinio juris is the key criteria, and one may doubt how suitable UN resolutions are to express it.
The Nuclear Weapons opinion notes that ‘a series of resolutions may show the gradual evolution of the opinio juris required for the establishment of a new rule’, 143 and stresses the importance of ‘the adoption each year by the General Assembly’ 144 of resolutions calling for the use of nuclear weapons to be prohibited.
The significance of repeating resolutions is unclear. Certain individual judges have found repetition to be important because the voting and passing of resolutions could be interpreted as state practice. 145 But the Nuclear Weapons opinion links repetition to opinio juris. Surely a legal conviction is not created by repeated expression. However, repetition may be relevant for determining evolutions within the opinio juris, for identifying increasing or decreasing support by various states: in short, for establishing the current opinio juris. 146 In this light, it seems more reasonable to interpret the Nuclear Weapons reference to repetition as a simple statement that opinio juris may evolve, without it being a criterion of any kind. It rather pertains to the modal effects of when a new opinio juris replaces the former. This happens whenever a new resolution expresses the opinio juris on a given subject.
All GA resolutions to which the Court has arguably attributed an impact on general international customary law are in fact, in their relevant provisions, declarations. This is illustrated by their titles – the Declaration on the Granting of Independence to Colonial Countries and Peoples (GA Resolution 1514 (XV) of 14 December 1960) in the Namibia and Western Sahara cases, the Declaration on Principles of International Law concerning Friendly Relations and Co-operation among States in accordance with the Charter of the United Nations (GA Resolution 2625 (XXV) of 24 October 1970) in the Western Sahara and Nicaragua cases, and the Declaration on the Prohibition of the Use of Nuclear and Thermo-nuclear Weapons (GA Resolution 1653 (XVI) of 24 November 1961) in the Nuclear Weapons case. This is hardly a coincidence, since it is precisely declarations that purport to deal with general international law, yet find no Charter authority for having any direct effect on it.
While declarations are the only adequate form of GA resolution to address general international law, and repetition rather pertains to modal effects, the three remaining criteria (conditions of the adoption, content and reality of the opinio juris) are, it is submitted, only various vehicles for the true test, which is intent. 147 Does a large majority of (the most concerned) states intend the resolution to express the current opinio juris?
Declarations in practice contain norms of lex ferenda which the Court has seen as a possible source of opinio juris. What then are the legal effects of these declarations?
The ICJ has not dealt with causative effect in the context of declarations. GA declarations may make general determinations of law, but not, by the nature of declarations, of specific facts or legal situations. 148 Their effect is then not causative but more akin to substantive. 149 If a declaration were nonetheless to make a determination of specific facts, there would be no basis for giving it effect since the GA has no power to apply general international law. Such determinations could therefore only trigger substantive effects in general international law if endorsed by a body competent to apply it (for example a state), in which case the legal effect does not flow from the UN body’s determination but rather from the later endorsement. Hence, a determination made in a GA or SC resolution cannot have any causative effect on general international law.
Substantive effects exist but are of a particular nature. Because the resolutions only inform the opinio juris, while the practice element of customary law is, in current ICJ jurisprudence, extraneous, the resolutions do not have any actual and autonomous substantive effects. Their effects are, one may say, pre-substantive, laying the ground for a real substantive effect if the missing element is provided. 150
The only finding concerning the modal effects attached to the above pre-substantive effects is that resolutions may show the evolution of the opinio juris. A new resolution that meets all the criteria immediately updates the opinio juris. Indeed, the pre-substantive effects on opinio juris can neither be deferred (current opinio juris cannot be constituted by state belief that a rule will be binding in the future), nor retroactive (the constitution of an opinio juris cannot imply that it already existed prior to that moment 151 ). Although an opinio juris cannot programme its own limitation in the future, it is by nature reversed by the appearance of a new opinio juris, and hence cannot be definitive (an opinio juris cannot be that states may not in the future have a new and contradictory opinio juris 152 ). Summing up, the effects are always immediate, non-retroactive and reversible. They are also murky in practice, but that does not affect the theoretical analysis. While the resolution shows when the pre-substantive effects start, it is impossible to know how long they will last, since the opinio juris may change under the impact of other influences. 153
There is a dichotomy in the legal effects given by the ICJ to GA and SC resolutions. The legal effects of recommendations and decisions are treaty-based, affect the UN legal order and include (i) the causative effect of factual and/or legal determinations; (ii) substantive (binding, authorizing and/or (dis)empowering) effects; and (iii) the modal effects that regulate when and how the substantive effects occur (as a rule, they are immediate, non-retroactive and reversible). The legal effects of declarations, which are based on customary international law and bear upon the international legal order, are more limited. Any determinations would be without causative effect, there are no direct substantive effects but only pre-substantive ones, and these are always immediate, non-retroactive and reversible.
The cause of this dichotomy is neither the type of resolution, nor the customary or conventional basis of the effects, nor the legal order affected. It is the difference between the intrinsic nature of the effects based on special powers and the extrinsic nature of the effects based directly on general international law. This holds true for each category of legal effects.
Determinations made in recommendations and decisions, unlike declarations, may have causative effect. This effect occurs when law is applied on the basis of a determination. Imagine that the ICJ were to acknowledge as an internal UN customary norm that, based on a sufficiently constant and uniform practice of recommendations combined with an opinio juris, the GA is not precluded from recommending measures to maintain peace and security when it determines that the SC is paralysed by a lack of unanimity among its permanent members. 154 Any future GA determination that the SC is so paralysed would have a causative effect on the internal customary norm. This is because the customary norm is within the GA’s power of application, and the effect is therefore intrinsic. No extrinsic causative effects are possible because general international law does not give the GA or SC any powers of application. Causative effect may only be intrinsic.
Recommendations and, especially, decisions may have substantive effects, while declarations only have pre-substantive effects. The effect operates at different stages of the normative process, either creating the substantive effect itself or merely laying the ground for its creation. UN resolutions can directly create special, usually UN, law, but can only have indirect effects on general international law by acting on one of the constitutive elements of customary law. Consider again the hypothesis discussed in the preceding paragraph – the resolutions given effect would be recommendations and would affect the UN legal order. Yet the (pre-)substantive effects would be similar to those encountered in the second section of this article – each recommendation would only contribute a part of the practice necessary, combined with an opinio juris, to constitute a substantive effect. This is because the effects would be extrinsic, based on the customary international law of the subsequent practice of the parties to a treaty. 155 On the other hand, if the ICJ were to grant substantive effects on general international law to an SC decision, 156 its effects would be similar to those encountered in the first section of this article. This is because the effects would be intrinsic, based on implicit powers in the UN Charter. Under the existing rules of formation of general international law, only intrinsic effects can be substantive, whereas extrinsic effects can only be pre-substantive.
The modal effects are also different. The modal effects of recommendations and decisions are flexible – immediate or deferred, retroactive or not, definitive or reversible/transient. The modal effects of declarations on opinio juris are inflexible – always immediate, non-retroactive and reversible (although murky in practice). In the first case, the adopting body has special powers capable of modulating the substantive effects – the modal effects are intrinsic. In the second case, it has no powers and the effects can only be extrinsic, directly based on the general international law of how customary law is created. This law allows for no modulation. Intrinsic modal effects are flexible; extrinsic legal effects are inflexible.
Let us sum up our conclusions:
| Effects . | Intrinsic . | Extrinsic . |
|---|---|---|
| Causative effect | Yes. | No. |
| Substantive effects: binding, authorizing or (dis)empowering. | Yes. In decisions; possibly some authorizing effects in recommendations. | Only pre-substantive. |
| Modal effects | Yes. Flexible. | Yes, but inflexible. |