1. Introduction
Questions of attribution to a state of internationally wrongful acts have frequently been at the heart of cases before international courts and tribunals. Reference can be made to the International Court of Justice's (ICJ or the Court) prominent Nicaragua case,Footnote 1 which concerned the responsibility of the United States for atrocities committed by the contras, the ICJ's Teheran Hostages judgmentFootnote 2 and the Tadić Appeals Chamber case,Footnote 3 decided by the International Criminal Tribunal for the former Yugoslavia (ICTY). Attribution issues also played a significant role in the ICJ's judgment of 26 February 2007 in the case concerning the Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herzegovina v. Serbia and Montenegro) (hereinafter Bosnia v. Serbia),Footnote 4 which will be the subject of this analysis. After the Court found that the massacres at Srebrenica committed by various groups of Bosnian Serbs constituted acts of genocide in violation of the Genocide Convention,Footnote 5 it dealt with the question of whether these acts could be attributed to Serbia and Montenegro (hereinafter Serbia).Footnote 6
Without anticipating the later analysis, one can say that the Court in several respects presented an overall conception of attribution which is not only in conflict with established principles of international law but also contrary to the needs of the international community. It is the purpose of this article to describe and comment on the conception followed by the ICJ. In this context reference will be made to earlier cases before international courts and tribunals dealing with questions of attribution as well as to the International Law Commission's (ILC) Articles on State ResponsibilityFootnote 7 (hereinafter ILC Articles) considered for the most part as reflecting customary international law.Footnote 8
After recalling the importance of attribution rules (section 2) and a brief summary of the relevant passages of the judgment in the Bosnia v. Serbia case regarding attribution (section 3), a closer look at various aspects of the decision will be taken (section 4). The article concludes by considering the future relevance of the ICJ's concept of attribution in the Bosnia v. Serbia case (section 5).
2. The relevance of attribution as a precondition of state responsibility
Rules of attribution are of considerable relevance in international law.Footnote 9 The reason for this lies in the simple fact that states as legal persons can only act through natural persons.Footnote 10 Without attribution the state is incapable of acting on the international plane. Therefore the ILC raises in the context of state responsibilityFootnote 11 for the purpose of determining whether a breach of international law has occurred, first, the question of attribution in order to define whether a certain act is an ‘(international) act of the State’, one which is of relevance under international law. Only in a second step is it asked whether this act of the state is contrary to international law and therefore an ‘internationally wrongful act of a State’. This order is emphasized not only by the ILC in Article 2 of the ILC Articles but also in the ICJ's judgment in the Teheran Hostages case.Footnote 12
The circumstances under which acts can be attributed to states have always been subject to intense discussion. What is agreed is that there is no easy rule that defines the requirements for attribution. This is because there is not a rule which would require a state's responsibility for all the acts committed on its territory, and nor is there a rule for all the acts committed by its nationals.Footnote 13 In order to determine the required link between the state and the acting natural person, different rules apply in different situations. The ILC has proposed altogether eight different attribution rules within Articles 4–11 of the ILC Articles. This analysis will demonstrate that there are different understandings of the content and function of some of these rules.
Attribution is accordingly an important element in the system of state responsibility. As such it forms part of this system's body of secondary rules, rules which are meant to apply to all situations in which primary rules providing for certain obligations are breached.Footnote 14
If the preconditions for an internationally wrongful act are met, the rules of state responsibility provide for two types of consequence according to the conception of the ILC: first, substantial consequences and, second, instrumental consequences.Footnote 15 Substantial consequences require the responsible state to make good the violation, which can be done by way of restitution or, if this is not possible, by way of compensation or satisfaction.Footnote 16 The instrumental consequences indicate that the victim state is also entitled to take measures in reaction to the violation, including such non-forcible actions which would otherwise be contrary to international law, that is, countermeasures.Footnote 17 This broad understanding of the consequences of state responsibility reflects the fact that the rules are meant to cover ‘all forms of the new relationship that may be established by international law by a State's wrongful act’.Footnote 18 This wide understanding of the consequences of state responsibility is convincing not only because it reflects the realities of international law but also because it is shared in international jurisprudence.Footnote 19 For considerations concerning attribution it is relevant in two respects. The attribution of an act to a state not only determines whether that state may have to grant a restitutio in integrum (restitution) or pay compensation; at the same time it determines whether the victim state may take action against the responsible state in order to enforce international law.Footnote 20 In considering the appropriateness of the rules of attribution one has to keep both aspects in mind, as will be seen.
3. The ICJ's concept of attribution in the Bosnia v. Serbia case
After an extensive discussion of the application of various articles of the Genocide Convention, the Court concluded that only the massacres at Srebrenica constituted violations of this Convention.Footnote 21 It proceeded with the question of whether the acts of those who committed the massacres were attributable to Serbia.Footnote 22 Only if this had been the case could the Court have found Serbia responsible for the massacres. The Court set out the procedure to be followed in this respect as follows:
First, it should be ascertained whether the acts committed at Srebrenica were perpetrated by organs of the Respondent, i.e., by persons or entities whose conduct is necessarily attributable to it, because they are in fact the instruments of its action. Next, if the preceding question is answered in the negative, it should be ascertained whether the acts in question were committed by persons who, while not organs of the Respondent, did nevertheless act on the instructions of, or under the direction or control of, the Respondent.Footnote 23
The ICJ followed this outline strictly and addressed first, with reference to Article 4 of the ILC Articles, the attribution rule concerning de jure organs.Footnote 24 As none of the persons or groups involved in the massacres at Srebrenica were found to have held the position of officially entitled organs under the internal law of what was at that time the Federal Republic of Yugoslavia (FRY), the Court denied attribution for actions of de jure organs.Footnote 25 Still under the heading of Article 4 of the ILC Articles, the Court proceeded, however, by raising the question of whether the acts could be attributed to Serbia as acts committed by de facto organs.Footnote 26 In this respect the Court, with reference to the Nicaragua case, applied a test of ‘complete dependence’.Footnote 27 According to the ICJ this test required proof that ‘the persons, groups or entities act in “complete dependence” on the State, of which they are ultimately merely the instrument’.Footnote 28 In applying this test to the facts of Bosnia v. Serbia the Court found that no such relationship existed between the FRY and the various examined groups of Bosnian Serbs.Footnote 29 This was mostly because ‘some qualified, but real, margin of independence’ was enjoyed by the Bosnian Serb leaders which signified no total dependence.Footnote 30
The Court then turned to Article 8 of the ILC Articles. Within its introductory explanation it pointed out as follows:
[T]he Court now addresses a completely separate issue: whether, in the specific circumstances surrounding the events at Srebrenica the perpetrators of genocide were acting on the Respondent's instructions, or under its direction or control. An affirmative answer to this question would in no way imply that the perpetrators should be characterized as organs of the FRY, or equated with such organs. It would only mean that the FRY's international responsibility would be incurred owing to the conduct of those of its own organs which gave the instructions or exercised the control resulting in the commission of acts in breach of its international obligations. In other words, it is no longer a question of ascertaining whether the persons who directly committed the genocide were acting as organs of the FRY, or could be equated with those organs – this question having been answered in the negative.Footnote 31
Article 8 of the ILC Articles, the ICJ found, laid down the applicable customary rule in this context. While the Court understood the notion ‘control’ within Article 8 of the ILC Articles to signify ‘effective control’, as provided for in the Nicaragua decision, it discussed whether the standard of ‘overall control’ applied by the ICTY in the Tadić Appeals Chamber case was preferable. In this respect the Court found that ‘logic does not require the same test to be adopted in resolving the two issues [whether an armed conflict is international and whether state responsibility is given] which are very different in nature’.Footnote 32 It furthermore criticized the overall control test as broadening the scope of state responsibility well beyond the fundamental principle that each states is only responsible for its own conduct – that is, the conduct of persons acting on its behalf.Footnote 33
In its application of Article 8 of the ILC Articles and in particular the effective control test, the Court came to the conclusion that it had been established neither that the Srebrenica massacres were committed on the instructions, or under the direction, of organs of the Respondent, nor that the Respondent exercised effective control over the operations in the course of which those massacres were perpetrated.Footnote 34
4. Analysis of the decision
In analysing the concept followed by the ICJ, three main aspects which are strongly related to one another and which overlap in part will be dealt with. In the first place, the distinction made by the Court between on the one side attribution of acts of de jure and de facto organs according to Article 4 of the ILC Articles and on the other side responsibility for acts committed by non-organs under the instruction and control of state organs according to Article 8 of the ILC Articles will be addressed (section 4.1). In analysing the ICJ's conception special emphasis will be laid here on a comparison with the concept of the ILC on which the Court relies. Second, the complete dependence test, which is according to the Court's perspective the sole existing attribution rule concerning de facto organs, will be addressed. Where the test possibly stems from will be analysed, and whether it is justified to regard it as the sole test of attribution for de facto organs (section 4.2). Here a closer look will be taken at the Court's own jurisprudence regarding similar situations. Lastly, whether the Court's approach meets the needs of the international community will be discussed (section 4.3).
4.1. The abolishing of Article 8 of the ILC Articles as an attribution rule, and its consequences
One key aspect of Bosnia v. Serbia concerns the legal effect ascribed to Article 8 of the ILC Articles. The Court, in distinction to the attribution rule for de facto organs (in its view the complete dependence test), did not regard Article 8 ILC Articles as an attribution rule at all. Considering the very open words by which the Court introduced its conception it seems rather surprising that this aspect of the judgment has seemingly not provoked other publicists' comments.Footnote 35 The Court introduced the relevance of Article 8 constellations as quoted above, the key passage being the following:
An affirmative answer to this question [that of whether the requirements of Art. 8 ILC Articles are fulfilled] would in no way imply that the perpetrators should be characterized as organs of the FRY, or equated with such organs. It would only mean that the FRY's international responsibility would be incurred owing to the conduct of those of its own organs which gave the instructions or exercised the control resulting in the commission of acts in breach of its international obligations.Footnote 36
What can be understood by the Court's expression in this judgment? The plain reading of the wording is that the ICJ is of the opinion that the responsibility incurred by the state under Article 8 of the ILC Articles flows from the conduct of the state's organs in giving the instructions or exercising the control in question, as opposed to the action of the instructed or controlled entities. Considering that the function of the attribution rules is to attribute to the state the conduct of persons who have acted against international law, the Court's refusal to consider the persons acting under such instructions or control as de facto organs, and its foundation of responsibility in Article 8 situations on the wrongfulness of the state organs' instructions or control, entirely stripped Article 8 of the ILC Articles of its character as an attribution rule. In particular, since this was held in such open language, one doubts whether the ICJ could have meant anything different from the plain reading of the judgment. The wording surprises in this respect, as on the one hand the Court uses the ILC Articles in structuring its analysis and as a basis for its examination of the problem in question, but, on the other hand, the Court in its judgment ascribed to Article 8 of the ILC Articles a function entirely contradictory to the ILC's understanding of this article.Footnote 37 Leaving aside the specific tests provided for in Article 8 of the ILC Articles, which will be addressed more closely below, Article 8 falls within the same Chapter II as Article 4, both categorized under the title ‘attribution of conduct to a state’. This is already an indicator of the effect the ILC intended to give to Article 8 of the ILC Articles. Further confirmation that the ILC understood Article 8 of the ILC Articles as a proper attribution rule is given by the commentaries to the ILC Articles.Footnote 38 Roberto Ago, the former Special Rapporteur of the ILC for the topic of state responsibility, regarded the preliminary versions of the former Article 8(a) of the 1996 ILC draft version which later turned into Article 8 of the 2001 ILC Articles as the attribution rule concerning de facto organs.Footnote 39 As the general concept and the understanding of the rules of Article 8(a) of the 1996 ILC Draft Articles were, apart from being specified, left unchanged during second reading, one can – contrary to the Court – see Article 8 of the ILC Articles regarding questions of terminology as an attribution rule for de facto organs. One can even regard Article 8 of the ILC Articles as one of the key attribution rules within the set of rules proposed by the ILC, as it is of a particular practical relevance considering that states in trying to camouflage their policies in one way or the other make use of private persons and groups. This is also the reason why Article 8 of the ILC Articles has met with special attention in legal writings.Footnote 40
Considering that in such writings the understanding of the ILC regarding Article 8 of the ILC Articles is unanimously supported,Footnote 41 it is completely unclear on what authority the Court relied in support of its understanding of the article. Although the Court makes reference to its former jurisprudence, the analysis below will demonstrate that the ICJ is in conflict with its own case law when it points out that the complete dependence test is the only test applicable to de facto organs. From the authors' perspective, the Court in Bosnia v. Serbia was accordingly the first not to ascribe to Article 8 of the ILC Articles the function of a proper attribution rule.
After this denial of the proper function of Article 8 of the ILC Articles, the way in which the ICJ then proceeds to apply this provision is equally remarkable. The Court seemingly interpreted Article 8 of the ILC Articles just like any other primary rule prohibiting certain acts of support granted by states to private persons or groups such as the indirect forms of violations of international law rules. Such forms found their expression, for example, in the Friendly Relations Declaration,Footnote 42 as the following example concerning indirect violations of Article 2(4) of the UN Charter shows: ‘Every State has the duty to refrain from organizing, instigating, assisting or participating in acts of civil strife or terrorist acts in another State’.Footnote 43
Regarding provisions of this kind a state cannot be held responsible for the person's act itself but only for its own state organ's support of such groups, just as in the same way the Court found that Article 8 of the ILC Articles provided only for responsibility concerning a state organ's instruction or control exercised concerning private actors. The main problem with this interpretation is that Article 8 of the ILC Articles – apart from being meant to belong to the body of secondary rulesFootnote 44 – can never be regarded as a primary rule since it is lacking the reference to a specific rule of international law. This shows that it was not meant to fulfil such a function. The lack of reference to a specific primary rule prohibition is obvious. Does it follow from this that the Court which relies on the responsibility of the instructing or controlling state organ considers that all prohibitions existing under international law without exception can also be carried out in these indirect ways? Even if this were the case, the question is why Article 8 of the ILC Articles should then still appear in the set of secondary rules on state responsibility. At least for the attribution of the state organ's action of instructing or exercising control it is not needed, since here Article 4 of the ILC Articles applies. There are accordingly a couple of questions left open in seeking to understand the Court's conception.
While one may for all these reasons be forgiven for thinking that the ICJ followed not just a new but also a rather inconsistent approach, another important question arises: are the differences between the understanding of the Court and that of the ILC only of a theoretical or dogmatic nature, or are there real practical differences arising from the different concepts? In other words, does it matter whether an act is directly attributable or whether responsibility is given based on wrongful behaviour in the context of such acts? In Bosnia v. Serbia the ICJ denied that the requirements for the application of Article 8 of the ILC Articles had been met. Even if the Court had regarded Article 8 of the ILC Articles as a proper attribution rule, it would have come to the same conclusion. For the present case applying a different view would, therefore, not have been of relevance. However, what if the Court had found Article 8 of the ILC Articles to be applicable according to its conception? Is it generally of any relevance if responsibility is based on instructions or exercised control and not on an attributable act?
The answer to this is in the affirmative; it does matter if one can hold a state directly responsible based on attribution or, as the ICJ states, only indirectly, by basing responsibility on acts like instructions or control. While there may not necessarily always be a difference regarding the consequences of attribution, important differences may, however, arise in three respects.
In the first place it makes a difference policy-wise whether the outcome of a legal analysis is that a state is the ultimate offender violating directly an international obligation or whether it is only held to have participated in the wrongful acts of somebody else. Had the requirements for Article 8 of the ILC Articles been fulfilled in the Bosnia v. Serbia case one would, following the ILC, have regarded Serbia as the ‘author’ of the massacres of Srebrenica, while the Court would have concluded that Serbia gave instructions or exercised control regarding a genocide committed by somebody else. There is a difference as to whether one can regard a state as a mass murderer or merely as an accessory to somebody else's mass murder. Had the Court in the Nicaragua case regarded the acts of the contras as attributable to the United States, the international community could have held the United States responsible for inter alia some of the gravest violations possible in the field of humanitarian law.Footnote 45 Instead the United States was with respect to the contras held responsible only for indirect violations of international law.Footnote 46
In the Teheran Hostages case the attribution of the continuing hostage-taking by the students, based on expressions of approval by, inter alia, the Ayatollah Khomeini, gave a new dimension to the ICJ's findings irrespective of the fact that Iran was also held responsible for the whole of the situation by reason of its omission properly to protect the US embassy and its personnel as required by the Vienna Convention on Diplomatic Relations.Footnote 47 The Court in this context pointed out that the result of this Iranian policy of demonstrating approval ‘was fundamentally to transform the legal nature of the situation’.Footnote 48 This shows that the Court itself demonstrated a consciousness that it matters whether one can attribute an act, even if responsibility for the consequences of the act can already be based on a certain failure to act. Attribution decides on the state's role as a mere accessory or as an author of the act. In other words it determines whether responsibility is engaged for an everyday peccadillo or a direct violation of international law.
Second, the events of 11 September 2001 have shown that questions of attribution are at the heart of discussions on reaction mechanisms with regard to certain dangers. In the field of self-defence it was and still is of key relevance whether one can regard the act of certain seemingly private groups as attributable to a state.Footnote 49 It is obvious that such an attribution is more difficult the more restrictive the attribution rules are. While according to the ILC a state's instruction or exercised control concerning violent acts of a certain group would justify considering that state itself as the author of an armed attack, the same result is not as easily achieved when no attribution is given. Considering the existence of an armed attack of a state based on that state organ's instruction or the exercise of control over somebody else's act is certainly not equally as easy.Footnote 50 Moreover, questions of proportionality may also arise as reactions to a state's armed attack will have to be different from reactions to the state's support for an armed attack by somebody else.
The same is true for countermeasures, which are equally subject to the special requirement of proportionality.Footnote 51 Can a victim state which is reacting merely to the instructing or controlling by another state of acts committed by a third party react in relation to that state in the same way as where these acts were attributable? Where the act is attributable to another state the victim can in principle react by suspending the same obligations as those violated which are owed to the other state. However, is the same true where the other state is not to be blamed for the immediate act but only for its participation in these? If there is no attribution this must also have consequences regarding the options of reaction.
Third, while the ‘instrumental consequences’ are therefore affected by the differentiation, the same is true for the ‘substantial’ ones – that is, restitution or compensation.Footnote 52 If an act is seen as attributable, the state is responsible for all the damage based on this act. Where the state is merely responsible for instructions or control exercised over somebody else's acts, the chain of causation is a different one.Footnote 53 The acts the state can be blamed for are more remote than otherwise would be the case and therefore not all the consequences of the ultimate act necessarily fall within the state's responsibility.
This shows that there can be differences following the two conceptions, which in certain cases may have an enormous relevance. The legal consequences ascribed to Article 8 of the ILC Articles is therefore not a marginal question.
Consequently, one can summarize that the Court, although referring to the ILC's conception, applied an understanding of Article 8 of the ILC Articles which is not in conformity with that of the ILC and which leaves many questions open. The differentiation between the concept of the ICJ and that of the ILC matters, since the respective concepts may produce different outcomes.
4.2. The complete dependence test (complete control test) as the only test concerning de facto organs
The second aspect which deserves a closer analysis is the Court's reliance on the ‘complete dependence test’, which it regarded as the sole test by way of which actions of de facto organs can be attributed to a state as such. In this respect the origin of the test will be analysed (section 4.2.1) as well as whether it can really be regarded as the sole test of attribution, in particular considering the Court's own jurisprudence (section 4.2.2).
The complete dependence test is sometimes also referred to as the agency testFootnote 54 or the complete control test,Footnote 55 which is unfortunate, as it leads to further confusion.Footnote 56 Therefore the term ‘complete dependence test’ as used by the ICJ is hereinafter followed.
4.2.1. The origin and practicability of the complete dependence test
In the Bosnia v. Serbia case the Court in applying the complete dependence test relied on its Nicaragua decision.Footnote 57 In the latter judgment the Court made the following statement:
What the Court has to determine at this point is whether or not the relationship of the contras to the United States Government was so much one of dependence on the one side and control on the other that it would be right to equate the contras, for legal purposes, with an organ of the United States Government, or as acting on behalf of that Government.Footnote 58
Attention was hardly ever paid to this passage as a separate test alongside that of effective control applied in paragraph 115 of the Nicaragua case. Seemingly, the idea of a separate standard of complete dependence was first prominently advocated in 1997, by Judge McDonald in her dissenting opinion in the Tadić Trial Chamber case, which in this respect found only limited support in legal writings.Footnote 59 Still, taking a closer look at a statement given by the Court later, in the Nicaragua case, concerning the attribution question, it seems fair to conclude that the ICJ in Bosnia v. Serbia worked on the assumption that complete dependence was meant to be an independent attribution test.Footnote 60 In paragraph 277 of the Nicaragua case the Court summarized its findings concerning the question of control over the contras as follows:
The Court is however not satisfied that the evidence available demonstrates the contras were ‘controlled’ by the United States when committing such acts. As the Court has indicated (paragraph 110 above), the extent of the control resulting from the financial dependence of the contras on the United States authorities cannot be established.
This shows that the Court regarded the considerations concerning complete dependence in paragraph 110 as a proper and independent attribution test.
However, one may wonder why – as indicated – this test was hardly ever regarded as a relevant attribution test. For example, this idea was neither supported by the majority in the Tadić Trial Chamber caseFootnote 61 nor followed by the Tribunal in the Tadić Appeals Chamber case. The latter worked on the assumption that the Nicaragua judgment established only one test,Footnote 62 that of effective control, which it criticizedFootnote 63 and therefore replaced with the overall control test.Footnote 64 The same is true for the ILC, which in its commentaries made it clear that the notion ‘control’ in Article 8 of the ILC Articles was to be understood as ‘effective control’.Footnote 65 In these commentaries on Article 8 it made no reference to a test of complete dependence, while quoting the passage in the Nicaragua case concerning effective control.Footnote 66 In legal writings also the effective control test is mentioned only when reference is made to the Nicaragua case.Footnote 67
One can only speculate about the reasons for ignoring the double-test approach of the ICJ in the Nicaragua case. One reason may well be that the wording of the judgment in Nicaragua concerning the attribution of the acts of the contras to the United States was interpreted to the effect that the effective control test was a sort of specification of the complete dependence test.Footnote 68
Another reason may be that the complete dependence test is from its conception a problematic test. The practical usefulness of the test must be doubted, considering that the evidentiary threshold of the complete dependence test is an exceptionally high one. Contrary to the opinion expressed by Judge McDonald,Footnote 69 one may furthermore have doubts that there are situations where the test is applicable without at the same time the effective control test applying. The practical relevance of the complete dependence test is therefore negligible.
There is another point of concern regarding the application of the complete dependence test. This can be made clear if one compares this test with the attribution rule for de jure organs. According to the latter a state is responsible for the persons it has officially designated to perform certain state functions. However, regarding de jure organs the state is not responsible for all of their acts, as they can still act within a private capacity. In principle, responsibility arises only for those acts taken in their respective capacities. Although this is contrary to the version of Article 5 of the 1996 ILC Draft Articles on State Responsibility not explicitly mentioned in Article 4 of the 2001 ILC Articles, this requirement is still necessary. This is not only reflected by the ILC's commentaries on Article 4 of the ILC Articles,Footnote 70 but also by Article 7 of the ILC Articles, which under certain albeit limited circumstances also provides for the state's responsibility for ultra vires acts of its de jure organs. Hence the state is not ipso facto responsible for all the acts of its de jure organs. The general relationship which exists based on a formal link is in itself not enough; it must be reflected in the action itself. The person must have acted within its capacity, which can only be examined focusing specifically on the act itself.
Regarding the complete dependence test the question arises as to whether there is also such a limitation regarding responsibility for private acts of a person. One publicist argues that, where the complete dependence test applies, ‘the non-state actor becomes a state organ de facto, and all of its acts, even those committed against explicit state instructions, would be attributable to the state’.Footnote 71 On the one hand this seems to be the logical consequence, considering that the test does not rely on any sort of accorded competencies; on the other hand this shows that it is unfortunate to base attribution on a general relationship while ignoring the relevance of that relationship regarding a specific act. The result is a test for de facto organs which is even stricter than the test for de jure organs, which can act in a private capacity without giving rise to the state's responsibility.Footnote 72 Determining attribution solely on the basis of the general relationship between the state and the person or group and ignoring the relevance of this link concerning the act in question is conceptually doubtful.
All this makes the complete dependence test one which can hardly be regarded as a practicable test leading to satisfactory results.
It is also interesting to note that the ICJ applies the complete dependence test as falling under Article 4 of the ILC Articles. On the one hand, the Court could not do otherwise, considering its understanding of Article 8 of the ILC Articles; on the other hand, one may have doubts that it fits into the conception of the ILC to apply Article 4 of the ILC Articles to such situations. Admittedly one has to concede that the ILC did not follow a too formalistic understanding of ‘organ’ as expressed in the commentaries to Article 4 of the ILC Articles, as follows:
[I]t is not sufficient to refer to internal law for the status of State organs. In some systems the status and functions of various entities are determined not only by law but also by practice, and reference exclusively to internal law would be misleading.Footnote 73
However, it is problematic to regard Article 4 of the ILC Articles as a sufficient basis also for attributing acts of groups one would regard prima facie as private groups. The intention of the ILC was rather to emphasize that the internal law need not always be sufficient to perform the task of classification.Footnote 74 The internal law's understanding of ‘organ’ may be a very peculiar or restrictive one.Footnote 75 For example, persons may perform public functions within a state based on competencies which have been granted not by way of a formalistic act but by way of a certain practice and use. Their actions are meant to be attributed. From an objective point of view they have to be regarded just like de jure organs which have been formally entitled. And as such they also share their denomination as de jure organs. Contrary to this, the complete dependence test focuses on persons which prima facie have nothing in common with de jure or such quasi-de jure organs, but whose acts are attributed on the basis of a different relationship. Such situations are not meant to fall under Article 4 of the ILC Articles, but under Article 8 of the ILC Articles. This is particularly true where the persons concerned do not act ‘within’ the legal system but outside it.
In this context it should be pointed out that disputes regarding the terminology are in principle of secondary relevance, as long as the same effect is attributed to the applied rules. However, considering the notion of de facto organs as being reserved for Article 4 and that of ‘agent’ for Article 8 of the ILC Articles – as is sometimes the case in the literatureFootnote 76 – can be criticized, as it is not useful to distinguish two situations which are so close to each other under different headings and rules.Footnote 77
Accordingly, although one may have strong doubts as to the complete dependence test's usefulness as a practically relevant attribution test, the Court in Bosnia v. Serbia was, however, correct in relying on the Nicaragua judgment as a basis for the complete dependence test. Doubts, however, arise regarding the application of this test under Article 4 of the ILC Articles.
4.2.2. Refutation of the complete dependence test as the sole test of attribution for de facto organs
Even if one therefore concluded that the complete dependence test is a test of attribution under international law, the question arises as to whether, as stated by the ICJ in the Bosnia v. Serbia case,Footnote 78 it is the only applicable test for the attribution of acts of de facto organs. This is the position taken by the Court, considering that it not only omitted to address other relevant tests for de facto organs but also degraded the tests provided for in Article 8 of the ILC Articles, namely the instruction test as well as the (effective) control test, to non-attribution tests.Footnote 79 Obviously this question is strongly linked to the discussion already mentioned concerning the abolishment as an attribution rule of Article 8 of the ILC Articles. It was stated above that in this respect the conception followed by the Court was in clear conflict with the ILC conception and also unanimous opinion in legal writings. In analysing the exclusiveness of the complete dependence test the focus of what follows will be on the Court's own jurisprudence in situations of a similar kind.
The ICJ's idea that complete dependence constituted the only attribution test for de facto organs has so far never been advocated elsewhere. As indicated above, in the interpretation of the Nicaragua judgment some experts relied on effective control as the only test applied in the Nicaragua caseFootnote 80 while others supported the idea of two independent tests.Footnote 81 Apart from this, other judgments as well as the ILC Articles indicated the relevance of other tests not concerned with the criterion of control, in particular the test of instruction. Accordingly, the Court followed a new approach.
Although in doing so the Court seemingly relied on its own jurisprudence, making reference to the Nicaragua case on several occasions,Footnote 82 the authors are of the opinion that the Court's concept is in conflict with its own earlier case law.Footnote 83 This is true, first, considering the instruction test as laid down in Article 8 of the ILC Articles. This test was not only applied by other international courts and tribunals,Footnote 84 it was also addressed by the ICJ in the Teheran Hostages case. Here, the Court had to deal with the question of whether the attack of the ‘militants’ on the US embassy engaged the direct responsibility of Iran. The Court described the relevant standard as follows:
Their conduct might be considered as itself directly imputable to the Iranian State only if it were established that, in fact, on the occasion in question the militants acted on behalf of the State, having been charged by some competent organ of the Iranian State to carry out a specific operation.Footnote 85
Although such attribution was denied on the facts of the case,Footnote 86 it becomes clear from this statement that the Court regarded the instruction test as an attribution rule.
Further to this Court's holding it is difficult to imagine a case where an attribution is more justified and required than in cases where an instruction by the state to a non-de jure organ is given to commit a certain act, for example the murder of a foreign state representative. The ICJ itself in the Bosnia v. Serbia case emphasized that the attribution rule for de facto organs is meant to prevent a state escaping international responsibility,Footnote 87 a view already expressed by many others on various occasions.Footnote 88 One cannot doubt that the instruction test is the clearest answer in fulfilling this demand.Footnote 89
While the instruction test was therefore acknowledged by the Court as an attribution rule, the same is true for at least the effective control test also laid down in Article 8 of the ILC Articles. The ICJ in the Nicaragua case regarded effective control as a genuine attribution test. To this effect reference can in the first place be made to the Court's elaborations in paragraph 115:
United States participation, even if preponderant or decisive, in the financing, organizing, training, supplying and equipping of the contras, the selection of military and paramilitary targets, and the planning of the whole of its operations, is still insufficient in itself, on the basis of the evidence in the possession of the Court, for the purpose of attributing to the United States the acts by the contras in the course of their military and paramilitary operations in Nicaragua.Footnote 90
For this purpose it would have been required ‘that [the] state had effective control of the military and paramilitary operations in the course of which the alleged violations were committed’.Footnote 91 These passages are proof that the Court in the Nicaragua case regarded the effective control test to be a genuine attribution test. This finding is furthermore supported in paragraph 277 of the Court's judgment in that case. Here the Court stated, following the considerations concerning complete dependence quoted above, as follows: ‘and it has not been able to conclude that the contras are subject to the United States to such an extent that any acts they have committed are imputable to that State (paragraph 115 above)’. The two tests of complete control and effective control may according to the Court in Nicaragua enjoy an independent character; however, both of these tests were meant to be proper attribution tests.
This is true even though the Court did not give a positive definition of what it understood by effective control. From the long elaborations on what did not suffice in this respect one may infer that the Court required that the state have a direct influence on the act or operation in question.Footnote 92 The state must be able to define its beginning and the way it is carried out, as well as its end.Footnote 93 An attribution under such circumstances is justified.
Another case in which the ICJ regarded the tests provided for in Article 8 of the ILC Articles as an attribution test is the comparatively recent judgment of the Court in Congo v. Uganda.Footnote 94 Although the Court did not here elaborate to any great extent on this topic, it becomes clear from the judgment that it regarded the tests embodied in Article 8 of the ILC Articles as proper attribution tests. After discussing the application of Articles 4, 5, and 8 of the ILC Articles,Footnote 95 the Court summarizes this analysis as follows: ‘the evidence does not suggest that the MLC's conduct is attributable to Uganda’.Footnote 96 Accordingly the tests reflected by Article 8 of the ILC Articles were given the same effect as those of Articles 4 and 5.
As indicated above, it was in part the reliance on the Teheran Hostages Footnote 97 and Nicaragua cases which led the ILC under the guidance of James Crawford to the current version of Article 8 of the ILC Articles as an attribution rule for de facto organs, and to draft it with explicit reference to the instruction test and the (effective) control test instead of referring simply to ‘in fact acting on behalf of that state’, as was provided for in the original draft elaborated under the direction of Roberto Ago.Footnote 98 Also Judge Simma, at that time chairman of the drafting committee concerned with state responsibility, worked on the assumption that the cases falling under Article 8 of the ILC Articles concerned cases of attribution.Footnote 99
It is accordingly not convincing that the Court in the Bosnia v. Serbia case considered the complete dependence test as the sole test for de facto organs. This is not only contrary to the well-established and widely accepted concept of the ILC as pointed out above, it is in particular not in accordance with the Court's own jurisprudence concerning the attribution of de facto organs. The approaches in the Teheran Hostages and Nicaragua cases are well founded, considering that it would seem rather too removed from reality not to regard a state which is initiating a murder by instructing the murderer or which is effectively controlling a massacre committed by a group of non-de jure organs as responsible based on an attribution.
4.3. The restrictive character of Article 8 of the ILC Articles as applied by the Court
A third aspect of the ICJ's judgment to be addressed is the restrictive character of the tests advocated by the Court in applying Article 8 of the ILC Articles. By rejecting Article 8 of the ILC Articles as an attribution rule the Court took a very restrictive perspective regarding the whole of the attribution concept. This restrictive approach is mirrored by regarding the complete dependence test as the sole existing attribution test for de facto organs – a test, which for reason of its required preconditions, is not likely ever to be applied in practice.
Although the Court followed – as discussed above – a very peculiar understanding of Article 8 of the ILC Articles, the way it applied the tests provided for therein is equally an expression of the Court's overall restrictive approach. In its judgment the ICJ discussed, in the context of the effective control requirement of Article 8 of the ILC Articles, whether the notion ‘control’ should not rather be understood as overall control instead of effective control. This idea of operating with the less restrictive standard of overall control introduced as mentioned above by the ICTY was dismissed by the Court for reasons to which one can in part easily subscribe. There is indeed no reason why the same rules need to be applied regarding issues which are so fundamentally different in nature.Footnote 100 However, the ICJ also reasoned that a move of the required standard from effective control to overall control would be broadening the scope of responsibility too far. Considering the requirements of the effective control test it must be regarded as a very restrictive test.Footnote 101
In this context it is regrettable that the Court did not address any of the new tendencies regarding attribution of acts of de facto organs which occurred after the events of 11 September 2001. After all, since the question was raised as to whether the acts of al-Qaeda could be attributed to Afghanistan, based on a so-called ‘safe-haven doctrine’ or acquiescence, a debate has been opened on the appropriateness of the more restrictive standards laid down in Article 8 of the ILC Articles.Footnote 102 In the Bosnia v. Serbia case the Court did not address this at all. All it addressed in this context was the idea of an overall control test to be applicable also in the field of state responsibility, a test which the ICJ found – as indicated – ‘unsuitable, for it stretches too far, almost to breaking point, the connection which must exist between the conduct of a State's organs and its international responsibility’.Footnote 103
One easy reason for not considering any of the other new ideas would be a silent application by the Court of the concept of intertemporal law.Footnote 104 The ICJ has to apply the law as it stands at the time when the dispute arises or the situation in question occurs and has accordingly to consider the wrongfulness of an act against the background of the law at the time when that act was committed. The Court relied – although in the unique way as discussed above – on the rules used in the Nicaragua case of 1986. If one takes a look at the period from the mid-1980s to the mid-1990s, when the massacre of Srebrenica took place, it is difficult to argue in favour of a change of the rule of effective control. This is also supported by the fact that the ILC in early 2001 presented its current version of the ILC Articles which in Article 8 also relied on effective control and which in December 2001 was commended to the attention of governments by the UN General Assembly.Footnote 105 Obviously one was at that time incapable of reacting to the new challenge which had just occurred three months before.
However, considering that the Court has frequently taken the opportunity progressively to develop the lawFootnote 106 against the background of the newly discussed tests, one might ask if such an opportunity was not missed regarding the attribution of acts of de facto organs. At least to elaborate on these new tendencies would have been useful, as the degree of uncertainty is enormous and there are still many questions, such as: have new attribution rules been accepted in customary international law? What is their possible character? Must the general rules as laid down in Article 8 of the ILC Articles be regarded as the universal ones which are, however, modified by a lex specialis in certain areas of international law?
From the Court's decision one could draw conclusions concerning at least some of these questions. The attribution rules are stricter than ever; states have to establish a relationship to the non-de jure organ which is almost similar to that of de jure organs. The standard of requirements for holding a state directly responsible for acts of non-de jure organs is accordingly extremely high, or, to put it in the words of the ICJ, ‘to equate persons or entities with State organs when they do not have that status under internal law must be exceptional’.Footnote 107 Further, regarding the question of lex specialis the Court pointed out that ‘[t]he rules for attributing alleged internationally wrongful conduct to a State do not vary with the nature of the wrongful act in question in the absence of a clearly expressed lex specialis’.Footnote 108
Still, one may have doubts whether these findings will at the same time fulfil the ICJ's target of trying to prevent states escaping responsibility.Footnote 109 Here lies another major weakness of the Court's overall conception. Is it not true that regarding attribution states can presently easily escape international responsibility? The Nicaragua case is an extreme example, where even massive forms of support, without which many of the atrocities committed by the contras would never have occurred, were regarded as insufficient for an attribution. This is remarkable, considering that the support was sufficient for the contras to increase in number from 500 to more than 12,000.Footnote 110 It seems obvious that one can find private groups to act for the cause of almost any interest which exists. States no longer need to act by way of their de jure organs if they wish to achieve certain aims; they can make use of the existing private groups. It is obvious that such groups can be as dangerous as states, as al-Qaeda has demonstrated. States do not need to control effectively or instruct these groups in order to have their aims fulfilled; considering that far too often the state's aims are shared with a private group certain forms of support will be enough. Considering the dangers flowing from private groups, the difficulties of fighting them under international law, and the abuse of such groups for political purposes by states, reflected by their enormous support for these groups, one might well raise the question of whether restrictive attribution rules are still justified. Rather, would a certain deterrent effect not be achieved by applying less restrictive rules? One may well discuss and even have doubts as to whether acquiescing in certain private acts (or a safe-haven doctrine) can in this context be regarded as a sufficient basis for attribution; however, if a substantial involvement of the state during the preparatory phase of an act or its execution can be proven, the state should not be able to escape direct responsibility for these acts.Footnote 111 In this way states would be prevented from conspiring with private groups for the purpose of violating other state's rights. Such rules have been overdue for quite some time. Considering that the judges of the ICJ have had before their eyes cases of the dark side of public–private partnership, such as the co-operation between the United States and the contras or the Taliban and al-Qaeda, it is regrettable that the Court did not see the international community's need for less restrictive attribution rules as well as the need to address this problem.
5. Conclusion and future prospects
From the above it can be concluded that without any apparent reason the Court applied a concept regarding attribution which in several aspects strongly deviates from commonly accepted concepts reflected in its own jurisprudence and the ILC Articles.
It is mostly its denial of Article 8 of the ILC Articles as an attribution rule which is the major point of concern. In arguing that responsibility in cases of control or instructions can only be based on the wrongfulness of the controlling or instructing de jure organs, the Court abolished a well-established concept and at the same time established an attribution standard which is even more restrictive than the approach followed by the Court in earlier cases, as well as that of the ILC. And instead of contributing to the certainty of law, the Court in this respect has clearly contributed to an increased uncertainty of law.
From the authors' perspective it is quite understandable that the Court in the light of such scenarios as that of 11 September 2001 did not consider attribution based on low-threshold requirements such as acquiescence, a safe-haven doctrine, or forms of substantive involvement. Apart from the fact that not all these tests would necessarily have been of relevance in Bosnia v. Serbia, it must be doubted – with substantial regret regarding an attribution based on substantive involvement – that they have acquired the status of universally applicable international law.Footnote 112 If at all they may play a role as lex specialis. However, considering the high degree of uncertainty of the law and the many divergent views found in legal writings, it would have been useful had the Court discussed attribution rules not only with regard to the Tadić Appeals Chamber case but also more generally. It is interesting to what degree the ‘conflict’Footnote 113 between the two institutions – the ICJ and the ICTY – concerning the required degree of control distracted the Court from the fact that both standards are in principle too restrictive and are not appropriate to meet the needs of the international community. One may accordingly regret that a chance to establish less restrictive standards, as may be due in today's world after the events of 11 September 2001, was missed.
Considering that the conception is at least in part confusing, it must be doubted that all members of the Court fully realized the eventual impact of the passages in the judgment concerning the effect of Article 8 of the ILC Articles. Considering also the special expertise of many of the members of the Court in the field of state responsibility, one may even be surprised that such an approach was followed.Footnote 114 That the Court's approach concerning attribution has not yet provoked strong criticism is seemingly due to the fact that it would not have made a difference regarding the outcome of the case had one applied the effective control test as a genuine attribution rule. However, while this is fortunately not the case, the remarkable restrictiveness of the Court's approach may have negative implications in future cases, although it must be doubted that the ICJ's perspective will prevail in the future. It seems to be a misapplication of established rules which is unlikely to find much support. One is even tempted to speak of a legal mistake made by the Court, one which it might be difficult to correct in the future.
It will be interesting to see how the Court will deal with its unfortunate approach should cases similar to Nicaragua or Bosnia v. Serbia occur again. The ICJ will probably find an opportunity to address the question of attribution in the pending proceedings of Croatia v. Serbia.Footnote 115
When considering questions of attribution there unfortunately seems still to be a lot of truth in the words written years ago by the Court's president, Rosalyn Higgins: ‘In the law of state responsibility one might be forgiven for thinking that there is almost nothing that is certain.’Footnote 116
