1. Introduction
One of the most challenging questions in the law of international organisations (IOs) concerns the apportioning of responsibility for internationally wrongful acts among IOs and member states of these organisations, which contribute troops to peace-support operations established or authorised by IOs.
For their operational activities, IOs, such as the United Nations or NATO, depend on member states placing organs (troop or police contingents) at the former's disposal. When wrongful acts are committed in the course of those activities, the question may arise as to whom those acts are to be attributed: to the organisation, to the contributing state, or to both of them jointly? The Draft Articles on the Responsibility of International Organizations for Internationally Wrongful Acts (DARIO), drafted by the International Law Commission (ILC), contemplate this scenario in the current Article 7 (2011 version).Footnote 1 Pursuant to this provision, attribution is a function of the level of effective control exercised by either the IO or the state.
It is the aim of this article to rationalise and defend the effective control standard as a means of apportioning responsibility in UN military operations. It is our view that only such a standard truly serves a preventative purpose (Section 1), and that, nonetheless, the UN's reservations as to the role of ‘effective control’ in respect of UN operations, as opposed to UN-authorised operations, are not entirely convincing (Section 2).
A second aim of this article is to review recent domestic and regional court decisions concerning the attribution of conduct in the context of UN peace-support operations (Section 3). The analysis starts with the well-known Behrami case decided by the European Court of Human Rights (ECtHR) in 2007, but continues with very recent cases from the UK and the ECtHR (Al Jedda, 2007–2011), the Netherlands (Srebrenica, 2008–2011), and Belgium (Mukeshimana, 2010). Whereas the court in Behrami failed to properly apply the effective control standard as laid down in Article 7 DARIO, the courts in the latter cases did seem to rely on ‘effective control’ to reach their conclusion (which, in all cases, was that the relevant acts were attributable not, or not only, to the organisation, but also to the state). At the inception of Article 7 by the ILC, there may have been scant evidence of this standard constituting customary international law, but Article 7 appears already to have had an effect on the practice of the courts. Section 4 concludes.
2. International Responsibility for Wrongful Acts Committed in UN Peace Operations: Article 6 v Article 7 Dario
In this article, we concentrate on issues of international responsibility arising in the course of peace-support operations (encompassing peacekeeping and peace-enforcement operations) conducted under the auspices of the UN. Internationally wrongful acts are most likely to arise in an operational context where UN-mandated actors (such as troops, police, and administrators) are in direct contact with populations. Such acts may specifically breach norms of international human rights law and international humanitarian law – norms of public international law which have the interests of the individual as the focal point of protection. Since peace operations are typically conducted in difficult circumstances – notably in conflict zones where there is a general breakdown of law and order – the commission of such breaches, even by UN personnel, is not imaginary. Examples include UN peacekeepers abusing local girls during peace operations,Footnote 2 or troops answerable to the UN arbitrarily arresting and detaining presumed criminals,Footnote 3 or even allegedly wrongfully targeting civilians during a military operation.Footnote 4 This may raise the question whether international human rights law or international humanitarian law applies at all to the UN, an international organisation which has not signed up to the relevant conventions. Our research question in this article is not, however, whether the UN is bound by substantive norms of human rights and humanitarian law, but whether violations of such norms committed by UN troops are necessarily attributable to the UN.
UN peace forces are normally considered to be subsidiary organs of the UN in the sense of Articles 7, 22, and 29 of the UN Charter, which authorise the UN in general, and the UN General Assembly and the UN Security Council in particular, to establish subsidiary organs as may be found necessary. In addition, semi-independently operating quasi-protectorate UN territorial administrations, such as the UN Mission in Kosovo (UNMIK), are organs of the UN.Footnote 5
In principle, all wrongful acts committed by UN personnel, acting as agents or organs of the UN, are attributable to the United Nations. Article 6 DARIO, which was largely inspired by relevant UN practice (see below for the UN's positions in this respect), and dicta of the International Court of Justice (ICJ) provideFootnote 6 that
[t]he conduct of an organ or agent of an international organisation in the performance of functions of that organ or agent shall be considered as an act of that organisation under international law whatever position the organ or agent holds in respect of the organisation.
It is our view, however, that this solution – which comes down to the wholesale attribution of acts of UN peace operations to the UN – may oversimplify the nature of UN peace operations. For every military operation conducted under UN auspices, UN member states place troops at the UN's disposal, which subsequently integrates them into its own subsidiary organs.Footnote 7 As a national force commander continues to exercise some measure of command over the national troops placed at the disposal of the UN, such troops are hybrids: they are both UN troops and national troops. This clearly complicates the responsibility question.
The ILC contemplated this scenario in Article 7 DARIO, which sets out the following rule relating to responsibility:
The conduct of an organ of a State or an organ or agent of an international organization that is placed at the disposal of another international organization shall be considered under international law an act of the latter organization if the organization exercises effective control over that conduct.
In the first commentary to Article 7 DARIO, the ILC clarifies that the article may have particular relevance for ‘the case of military contingents that a State places at the disposal of the United Nations for a peacekeeping operation’, where ‘the State retains disciplinary powers and criminal jurisdiction over the members of the national contingent’. It indicates that Article 6 DARIO (relating to the attribution of acts of agents and organs of the IO to the IO itself) is not applicable to such cases of secondment, but only to cases of full secondment where the organ of the seconding state becomes an organ of the IO. In case of full secondment, only the responsibility of the organisation can be engaged.
Pursuant to Article 7 DARIO, the burden of responsibility lies on whoever is exercising effective control over the military contingents: the IO (the UN in our case), the contributing state, or both of them jointly (shared responsibility). It flows from the standard set out in Article 7 that contributing states can, in specific circumstances, be held liable for acts carried out within the framework of IOs. To the extent that contributing states, within an operational framework supervised by IOs or their organs, exercise effective control over acts harming third parties, their responsibility can, and should, be engaged.
The rule laid down in Article 7 deserves full support, as it creates a proper responsibility regime that locates responsibility with the actor who is in a position of control over the wrongful acts giving rise to responsibility. This allows the actor to take measures to prevent the commission of the wrongful act. Eventually, such a regime deters the commission of wrongful acts, and furthers compliance with the law. Deterrence and compliance are not particularly furthered if the actor who exercises effective control over the operations of the troops is not exposed to responsibility: such an actor can take a free ride on another actor who is unjustly held responsible.Footnote 8 Ultimately, only a principle of attribution which locates responsibility with the actor who directly ordered, or could have prevented the violation of certain rights, appears to be sound. Arguably, this is the rationale behind the ‘effective control’ basis for attribution (to which we refer in the next section). If, as a result of the command structure of a UN operation, sufficient discretion is granted to contributing states for them to be in control of acts of the troops which they contributed to the operation, responsibility should lie with the contributing states.Footnote 9 If, in contrast, the UN keeps contributing state troops on a tight leash, by severely limiting the powers of the national force commander and strengthening the powers of the UN command, it is the UN that exercises effective control, and it is the UN whose responsibility will be engaged. If, because of circumstances, both the UN and the contributing state exercise joint, parallel, and/or roughly equal control, the responsibility of both will be engaged jointly and severally. Co-responsibility may in fact be the more typical situation in UN peace operations.Footnote 10 The existence of co-responsibility does not mean that states cannot be sued and be held responsible separately, however, as the court cases discussed in the next section demonstrate.
The argument that ties liability to effective control and the capacity to prevent rights violations in the context of UN peace operations has been made most convincingly by Dannenbaum.Footnote 11 This is a taxonomy which reflects the effective control criterion and therefore deserves support. Dannenbaum distinguishes five different factual patterns which attract liability:
1. Ultra vires acts of troop-contributing states, which the UN cannot prevent and, therefore, for which the contributing state incurs liability.Footnote 12
2. Acts carried out pursuant to authorised discretion by both the UN commander and the national force commander (which means that the respective commanders have been given a choice to carry out, or not carry out, a certain act), for which shared liability may apply because the two, or alternatively one of them, could have prevented the act, depending on the level of control exercised. In such a case liability will apply to the commander in a position to prevent.Footnote 13
3. UN orders instructing international crimes to be committed, which are obviously attributable to the UN, but also to the contributing state, which has a duty to disobey manifestly unlawful orders (both the UN and the contributing states are responsible in this scenario).Footnote 14
4. UN orders instructing violations of human rights law and humanitarian law which do not rise to the level of international crimes, which are only attributable to the UN, and not to the contributing state, which has no right to disobey orders in the chain of command (only the UN is responsible in this scenario).
5. Omissions, representing failures on the part of the UN to provide sufficient support, as a result of unwillingness or inability on the part of the UN (for example, lack of resources, readiness), and that are therefore attributable to the UN.
As far as the fifth scenario is concerned, it is noted that if the UN merely obstructs action or fails to exercise sufficient oversight without making action impossible, the contributing state retains some policy discretion, as a result of which responsibility may reside with both the UN and the contributing state.Footnote 15
The bottom line of this argument is that attribution is not based on the fact that troops wear a blue helmet – in which case all acts would be attributed to the UN – but on the level of effective control exercised – in which case acts are attributed to either the UN or the contributing state, or to both jointly. This is not too revolutionary an idea, as it was already implied by Seyersted in his 1966 monographFootnote 16 and, as noted above, has recently been laid down in Article 7 DARIO.Footnote 17
In its commentary to Article 7, the ILC gives some specification of the criterion of control. It states that the criterion is based on ‘factual control that is exercised over the specific conduct taken by the organ or agent placed at the receiving organization's disposal’,Footnote 18 and suggests that one should inquire who exercises ‘effective control’ over the troops.Footnote 19 The ILC remains rather vague as to what ‘effective control’ would look like in practice, limiting itself to agreeing with the UK position that regard should be had to the ‘full factual circumstances and particular context’.Footnote 20 In UN operations, contributing states typically retain full and exclusive strategic level command and control over their personnel and equipment, whereas the UN exercises operational authority over the troops.Footnote 21 In such a situation, the UN and the contributing states could be considered to both exercise effective control over the troops, and their responsibility for wrongful acts will be jointly engaged. Nonetheless, in emergency situations, or on the basis of specific arrangements, effective control may shift to either the UN or the contributing state, as a result of which the UN or the contributing state's sole responsibility will be engaged.
The facts underlying the Srebrenica and Mukeshimana court decisions, which are discussed below, are examples of emergency situations in which regular command structures have broken down, with attendant consequences in terms of apportioning responsibility. An example of a specific arrangement for the exercise of command and control is offered by the joint operation between the UN military mission in Côte d'Ivoire (UNOCI) and the French mission ‘Force Licorne’. Pursuant to the relevant UN Security Council resolutions, Force Licorne was authorised to use all necessary means in order to support UNOCI in accordance with an agreement to be reached between UNOCI and the French authorities.Footnote 22 This agreement has not been made publicly available, but it has been reported that the arrangements between the UN and Licorne were to include the provision by Licorne of a guaranteed Quick Reaction Force (QRF) in support of the UN Force Commander, which, upon deployment, was to report under the Tactical Command of the UN (sector) commander in whose area it operated.Footnote 23Arguably, any breaches of international law that were committed by this French QRF, while it was placed at the disposal and under the command of the UN, are attributable to the UN on the basis of Article 6 or 7 DARIO. Conversely, insofar as Licorne forces were not placed at the UN's disposal, or at least remained under French command, including operational command, their wrongful acts should be attributed to France rather than to the UN, on the basis of Article 7.Footnote 24
The ‘effective control’ standard requires that responsibility be delimited on the basis of a detailed factual inquiry. Formal arrangements may provide guidance but are not necessarily conclusive. Ultimately, the reality of operational command and control on the ground should be decisive for a responsibility regime to serve a preventative function. Even in respect of a UN operation proper, in fact, this reality may, as Srebrenica and Mukeshimana remind us, also be that control is wrested, possibly in its entirety, from the UN and transferred to the troop-contributing states, whose responsibility should then logically be engaged.
Importantly, by endorsing the effective control standard, the ILC has distanced itself from the ECtHR 2007 Behrami decision,Footnote 25 in which the Court held that UN Security Council (UNSC) Resolution 1244, which set up UNMIK, gave rise to a chain of command under which the UN Security Council ‘was to retain ultimate authority and control over the security mission’. According to the Court, the Security Council only ‘delegated to NATO (in consultation with non-NATO member states) the power to establish, as well as the operational command of, the international presence’.Footnote 26 Consequently, the troop-contributing NATO and UN member states could not be held responsible for purported rights violations. The ‘ultimate control and authority’ standard espoused by the ECtHR in Behrami was based on neither Article 6 nor Article 7 DARIO. As far as Article 7 is concerned, the Court did not inquire whether or not the UN exercised effective control but satisfied itself with ascertaining that the UN, at least on paper, exercised ultimate control. As far as Article 6 is concerned, the Court did not hold that the troop-contributing states acted as organs of the UN. Put differently, the Behrami standard has no basis in the DARIO – nor, it seemed, in applicable customary international law.Footnote 27
While the ILC did not seem to agree with the decision in Behrami, it did agree with a judgment rendered one year later by a Dutch District Court, which attributed the Dutch UN troops’ failure to prevent the killing of a number of Bosnian men in the Srebrenica genocide to the UN rather than to the Netherlands as the troop-contributing state, considering it ‘to be in line with the way in which the criterion of effective control was intended’.Footnote 28 In the Dutch Srebrenica case, victims had brought suit in the Hague District Court (a first instance court) against both the UN and the Netherlands as the UN member state contributing the troops for the UN peace mission in Bosnia Herzegovina. The petitioners were twice rebuffed by the Court in 2008, which ruled in one case that ‘actions of a contingent of troops made available to the United Nations for the benefit of the UNPROFOR mission … should be attributed strictly, as a matter of principle, to the United Nations’,Footnote 29 and in another that the UN was entitled to immunity.Footnote 30 In 2011, however, the Court of Appeal in The Hague surprisingly overruled the first instance decision in relation to the international responsibility question, and held that the acts of the Dutch UN mission were attributable to the Dutch government on the basis of the ‘effective control’ standard laid down in Article 7 DARIO. This momentous decision is discussed in more detail in Section 4.1.
What is interesting about the first instance Srebrenica decision, however, is that it is clearly informed by the UN Secretary General's (UNSG) official position on the question of apportionment of responsibility in UN peace-support operations. The UNSG's position makes Article 7 DARIO seem almost entirely redundant as it considers UN peacekeeping contingents to be UN organs pursuant to Article 6 DARIO. According to the UNSG, if UN operations are conducted under UN command and control (the classic peacekeeping mission), they are transformed into subsidiary organs of the UN in the sense of Article 6, and the UN assumes all responsibility. Alternatively, if the operation is merely authorised by the UN but conducted under national or regional command (such as the operation in Libya authorised by UN Security Council Resolution 1973(2011)), all responsibility lies with the state or the regional organisation, and the UN assumes no responsibility whatsoever. As the UN Office of Legal Affairs stated in 2011:
It has been the long-established position of the United Nations, however, that forces placed at the disposal of the United Nations are ‘transformed’ into a UN subsidiary organ and, as such, entail the responsibility of the Organization, just like any other subsidiary organ, regardless of whether the control exercised over all aspects of the operation was, in fact, ‘effective’. In the practice of the United Nations, therefore, the test of ‘effective control’ within the meaning of [Article 7] has never been used to determine the division of responsibilities for damage caused in the course of any given operation between the United Nations and any of its troop-contributing states. This position continued to obtain even in cases – like UNOSOM II in Somalia – where UN command and control structure had broken down.Footnote 31
In this binary model, there is no room for an inquiry into who was actually exercising effective control over the conduct of the mission. If the effective control standard does not applyFootnote 32 it is indeed immaterial, as the UN observed in its comments, whether the UN's command and control structure had broken down or whether the troop-contributing state retained residual control in matters of disciplinary and criminal prosecution, salaries and promotion for the duration of the operation;Footnote 33 what matters is simply whether the operation is a UN operation or a UN-authorised operation.
3. Assessing The Position of the United Nations
As set out at the end of the previous section, the UN insists on exercising exclusive effective control over UN peace operations. In many cases, however, the assumption of exclusive UN control regarding the organisational legal status of a UN peace operation has no foothold in reality. As Sari and Dannenbaum point out, in practice the UN has full command or operational control over national contingents only very rarely. These contingents retain, as state organs, a strong relationship with the state, which continues to exercise effective command (through the national force commander) and control.Footnote 34 This has obvious consequences for the liability locus, at least if one accepts that effective control determines with which actor liability lies. Indeed, if the national contingents continue to answer to the state, one cannot reasonably claim that peacekeeping missions are merely subsidiary organs of the United Nations under Article 6 DARIO, as the UNSG claims.
Somewhat confusingly, perhaps, the UN itself, in various statements, seems to have accepted that it does not always exercise effective operational control over UN peace operations. Thus, it has left the door open for contributing state responsibility within the framework of UN peace operations pursuant to Article 7 rather than Article 6 DARIO. In an early comment to the DARIO, the UN stated that ‘[a]s a subsidiary organ of the United Nations, an act of a peacekeeping force is, in principle, imputable to the Organization’Footnote 35 – a statement which implies that there could be exceptions to the rule. The existence of the effective control standard as an appropriate yardstick to apportion liability between the UN and its contributing states is even further strengthened by the UN's statementFootnote 36 that
[w]hile the submission of [periodic] reports provides the [Security] Council with an important ‘oversight tool’, the Council itself or the United Nations as a whole cannot be held responsible for an unlawful act by the State conducting the operation, for the ultimate test of responsibility remains effective command and control.
This undercuts the rationale of Behrami, which considered the submission of reports to the Security Council as decisive for control and responsibility purposes. And tellingly, in its 2011 comments,Footnote 37 the UN explicitly rejected the Behrami ‘ultimate authority and control’ standard in favour of an ‘effective command and control’ standard:
The recent jurisprudence of the European Court of Human Rights, beginning with the Behrami and Saramati case disregarded this fundamental distinction between the two kinds of operation [UN operations and UN-authorized operations] for purposes of attribution. In attributing to the United Nations acts of a UN-authorized operation (KFOR) conducted under regional command and control, solely on the grounds that the Security Council had ‘delegated’ its powers to the said operation and had ‘ultimate authority and control’ over it, the Court disregarded the test of ‘effective command and control’ which for over six decades has guided the United Nations and member States in matters of attribution.
In truth, this qualification may have to be understood in the context of ultra vires acts (although it is noted that Article 8 DARIO attributes ultra vires acts of organs or agents of the organisation – which peacekeeping contingents are in the view of the UN – to the organisation itself).Footnote 38 In addition, as set out above, the UN has clarified that ‘this test of “effective command and control” applies “horizontally” to distinguish between a UN operation conducted under UN command and control and a UN-authorized operation conducted under national or regional command and control’, and not ‘vertically’ in the relations between the United Nations and its troop-contributing states.Footnote 39 But it is conspicuous that the UN Secretariat terminates its discussion of Article 7 DARIO by supporting the latter's inclusion – and thus the existence of an effective control standard in all cases in which member states place troops at the UN's disposal – as ‘a general guiding principle in the determination of responsibilities between the United Nations and its member States with respect to organs or agents placed at the disposal of the Organization’.Footnote 40 In so doing, the UN seems to backpedal on its embrace of the binary model of UN operations versus UN-authorised operations. It does not exclude that also in UN operations, responsibility could, at least under certain circumstances, shift to contributing states if the latter exercise effective control. This will presumably happen when those member states back out of the UN's command structures by giving instructions that contradict UN instructions.
Finally, even if one is of the view, as the present author is, that national troop contingents are not to be considered as organs of the UN as they are not fully seconded to the UN (the contributing state retaining a measure of command and control), the question remains whether they may not qualify as agents of the organisation pursuant to Article 6 rather than as organs of a state pursuant to Article 7. An ‘agent’ of an organisation is an entity which is charged by the organisation with carrying out, or helping to carry out, one of its functions, and thus through whom the organisation acts.Footnote 41 For purposes of attribution, agents of IOs are treated on the same footing as organs of IOs: both engage the responsibility of the IO by their wrongful conduct.
It has been our position that national troop contingents taking part in UN operations are organs of a state that are placed at the disposal of the UN, as a result of which the legal regime of Article 7 rather than Article 6 controls this issue. But still, the concept of an ‘agent of an organisation’ may be so malleable as to possibly include national troops that carry out a peace and security-related mandate of the UN. The UN, however, seems to have excluded this interpretation: criticising the ILC's broad concept of ‘agent’, it even pointed out that it does not necessarily regard persons and entities who perform functions that are also performed by the organisation as ‘agents’ of the organisation, ‘but rather as partners who assist the Organization in achieving a common goal’.Footnote 42 The UN's responsibility would then not be engaged by the conduct of mere ‘partners’. Remarkably, indeed, the UN does not consider the performance of mandated functions as ‘conclusive’, but rather draws attention to various factors that should be considered on a case-by-case basis, ‘such as the status of the person or entity, the relationship and the degree of control that exists between the Organization and any such person or entity’.Footnote 43 Apparently, in the UN's view, the ‘control standard’ is also relevant in the context of Article 6. Ultimately, it seems, an IO's responsibility is simply engaged by the acts of persons or entities who had a sufficiently close relationship with the IO or over whom the IO had sufficiently strong (effective) control. Both Articles 6 and 7 DARIO then appear to be manifestations of one and the same rule.
Having analysed the UN's official statements on control over UN peace-support personnel, it is now apt to examine what might explain the insistence of the UN that it exercises effective control. At first sight, it may seem strange that an actor is willing to shoulder the burden of exclusive responsibility – it is indeed much more common for actors to pass the buck and come up with all sorts of arguments so as not to be held responsible. On closer inspection, the political rationale of the UN's position is rather self-evident. Claiming that the UN and a contributing state are jointly responsible for a wrongful act, or even that the contributing state is exclusively responsible, would amount to admitting that the UN does not and cannot exercise effective control over its own operations.Footnote 44 This is not exactly an image that a dynamic international organisation wishes to project, and partly explains the UN's observation in its 2011 comments to the ILC that ‘[f]or a number of reasons, notably political, the UN practice of maintaining the principle of UN responsibility … is likely to continue’.Footnote 45
Incidentally, it is relatively harmless for an IO to attract responsibility for itself, as there are no, or very few, mechanisms to hold such organisations to account.Footnote 46 The Dutch Srebrenica case aptly illustrates how victims are then caught between the devil and the deep blue sea. On the one hand, the Court of Appeal, hearing the case against the UN, held that the UN enjoyed immunity in proceedings before Dutch courts on the basis of Article 105 of the UN Charter and the UN Convention on Privileges and Immunities. On the other hand, the Court tersely noted that victims had access to reasonably available alternative remedies against the Dutch state – the contributing state which could be, and was, also sued by the victims,Footnote 47 while ironically the first instance court had earlier held in the case against the Dutch state that the acts of a contributing state's personnel – in line with long-standing UN practice – were attributable to the UN and not to the contributing state.
Another important political rationale underlying the view of the UN on the question of the attribution of wrongful conduct to either the UN or its contributing states is the need to safeguard the capability of the UN to send peace missions to far-flung corners of the world. The argument might be made that if states incur responsibility when they contribute troops to the UN, they might well refrain from contributing troops to future UN operations. Pursuant to this utilitarian approach, a liability regime should not be pushed to such limits that the very operation of peacekeeping missions is compromised. Such an approach is consequentialist and extrinsic in that it emphasises the end-goal and the possible spillover effects of an effective control-based liability regime (it factors in the adverse repercussions on member states’ willingness to contribute troops) rather than the intrinsic fairness of the regime.
It is interesting to observe in this respect that the courts’ quasi-automatic attribution of peacekeepers’ conduct to the UN rather than to the troop-contributing nations coincides with the post-1999 international ‘Responsibility to Protect’ discourse. This discourse emphasises the need for international intervention in conflict-prone regions where vulnerable populations are at risk, and thus the need to increase the UN's peacekeeping commitments. Arguably, the very effectiveness of such commitments may suffer when member states refuse to provide troops for fear of being held responsible for the wrongful acts of troops under their control. Of course, such responsibility disappears if contributing states cede effective control entirely to the UN, but given the historical impossibility of creating a standing UN armyFootnote 48 it is not to be expected that contributing states stand ready to transfer control over their own troops to the UN, however beneficial this could be for the effectiveness of UN missions. This leaves us only with a possible refusal by contributing states to commit sufficient troops. But is there indeed a genuine risk that contributing states will refuse to put troops at the UN's disposal if the effective control standard is embraced? Dannenbaum, while noting that ‘states participate in peacekeeping for a variety of complex reasons’Footnote 49 – of which the risk of incurring costs as a result of potential liability is but one factor – appears ready to countenance the reduction of contributions if ‘in exchange [this reduction] would improve the quality of those that are deployed’.Footnote 50 Granted, the risk of liability may encourage states to train troops with a view to preventing rights violations, so that the troops that are on the ground are model troops. But is there no tipping point: may the adaption costs not outweigh the perceived benefits of participation in peacekeeping missions, possibly rendering states more reluctant to commit troops in the first place? And if states continue to contribute troops, would they not tend to micro-manage their troops, thereby undermining the UN command structure?
Some court decisions appear to be informed by such concerns, notably the ECtHR's decision in Behrami (2007) and the Dutch courts’ decisions in Srebrenica (2008, 2010).Footnote 51 It is, however, conspicuous that no government has taken issue with the ILC's effective control standard in its comments to the ILC on the grounds that it has adverse political effects. In the latest round of comments (2011), only three states (Austria, Belgium, Germany) gave input regarding Article 7 DARIO, and none of them criticised the content of the provision, rather on the contrary.Footnote 52 Earlier, states such as Denmark and Poland had already rejected the automatic exclusive attribution of acts performed during UN peace operations, and emphasised the importance of an effective control standard in apportioning responsibility.Footnote 53
These governmental positions may surprise, as they attract member state responsibility for wrongful acts committed by troops seconded to the UN. At the same time, they are understandable, as refusing to accept responsibility would amount to accepting that the troop-contributing state has no control over the troops that it places at the UN's disposal – a position that, for reasons of state sovereignty and democratic control over armed forces, is not desirable (this is the mirror image of the UN's position). Also, as hinted at above, states have always retained a measure of control over and, even more, have micro-managed the troops they second to the UN, a situation which has been tolerated by the UN as long as the command structure is not impeded.Footnote 54 In fact, the very willingness of states to contribute troops to the UN hinges on their retaining control. Findings of responsibility of contributing states reflect the existence of such control, and ultimately reinforce it. Accordingly – and somewhat counter-intuitively perhaps – states may be more, rather than less willing to contribute troops to the UN if responsibility were to be based on effective control.
There is an obvious danger that the enhanced control exercised by states may prove to be an impediment to the fulfilment of the mandate of the UN operation, and thus compromises the political goals pursued by the UN. The UN, however, should counter this danger by urging states to subscribe to the UN command structure, and attract responsibility for itself – which indeed it has (these are the ‘political reasons’ referred to by the UN in its response to the ILC). In fact, this tug of war between the UN and the contributing states can be found in almost all UN operationsFootnote 55 and, as observed above, will typically result in the contributing states retaining full and exclusive strategic level command and control over their personnel and equipment, and the UN exercising operational authority.Footnote 56 Both contributing states and the UN then exercise effective control over the troops, and share responsibility for any wrongful acts that may be committed. Such shared responsibility need not necessarily be joint responsibility, but could well be separate responsibility for violations of different obligations albeit in the context of the same act. For instance, the contributing state's responsibility for commission may be engaged (because, for example, it gave orders to commit violations), whereas, arising out of the same act, the UN's responsibility for negligence may be engaged (for example, because the UN failed to prevent the commission of the violations, although it had the power to do so).
A regime of shared responsibility is desirable since, as Leck also pointed out, it prevents the ‘behavioural externalities and other undesirable consequences’ that flow from attributing wrongful acts to either the UN or the contributing states.Footnote 57 If those acts are attributed solely to the UN (that is, the logical consequence of the Behrami ‘ultimate authority and control’ standard), the contributing states may be tempted to take a free ride without the UN having the power and control to correct missteps, or – as Bell has noted – ‘recipient states of peace support operations [may] vigorously oppose the entrance of UN missions because of the fear of impunity for any resultant damages’,Footnote 58 or the UN may even hesitate to deploy missions in the first place if it attracts all responsibility.Footnote 59 If, conversely, wrongful acts are attributed solely to the contributing state, the result could be state micro-management of UN operations or, possibly, reluctance by states to contribute troops. Both scenarios are undesirable. A rule of shared responsibility is preferable instead, as it locates responsibility where it is due – where effective control is exercised, typically by both the UN and the contributing states – and ultimately limits the separate responsibility of the actors involved. This rule does justice to the political autonomy of both the UN and the contributing states in designing and contributing to UN peace operations, and is arguably the only attribution rule that reconciles rival political considerations as well as closes any impunity gaps.Footnote 60
The responsibility regime discussed above refers to the external responsibility of the actor who wronged the injured party. The externally responsible actor (the UN and/or the contributing state) may eventually, however, not necessarily have to provide the full amount of reparation due to the injured party: the UN and the contributing state may decide inter partes on the apportionment of the reparations between them. However, such an internal arrangement does not bind third parties.
The UN, for instance, appears as one front vis-à-vis the injured party, in the sense that the latter could only engage the responsibility of the UN and not of the contributing state – as noted above, the UN considers the troops of the contributing state to be merely subsidiary organs of the UN pursuant to Article 6 DARIO. However, it has not excluded that, under certain circumstances, it may want to revert to the contributing state. Citing the Model Memorandum of Understanding between the UN and troop-contributing states, the UNSG observedFootnote 61 in its latest reflections on Article 7 DARIO:
Mindful of the realities of peacekeeping operations but keen to maintain the integrity of the UN operation vis-à-vis third parties, the United Nations has struck a balance, whereby it remains responsible vis-à-vis third parties, but reserves the right in cases of gross negligence or wilful misconduct to revert to the lending State.
In stating that a contributing state and an IO could conclude an agreement with the receiving IO over who incurs responsibility when an organ or agent is placed at the disposal of the latter organisation, the UN, however, appears to have distanced itself from the ILC: the ILC noted that such an agreement ‘is not conclusive because it governs only the relations between the contributing State or organization and the receiving organization and could thus not have the effect of depriving a third party of any right that that party may have towards the state or organization which is responsible under the general rules’.Footnote 62
States and IOs may indeed be free to decide, inter partes, on the ultimate apportionment of the payment of damages on the basis of an agreement as contemplated by Article 9 of the Model Memorandum of Understanding between the United Nations and Participating States contributing resources to a UN Peacekeeping Operation. Such an agreement could allow one actor to recover from another, whether or not the other is also liable under international law vis-à-vis the third party.Footnote 63 However, such an agreement cannot be detrimental to third parties. Representing the principle of res inter alios acta, it does not diminish the right of third parties to claim the entire amount from the party that is responsible (whether or not jointly). And precisely because victims can often more easily file complaints against states than against IOs, one should be cautious in recognising an IO's unilateral acceptance of responsibility as conclusive: it may serve to exclude the responsibility of the contributing state and adversely affect victims’ interests. In the final analysis, all responsibility should rest on the intensity of control exercised by either the contributing state or the UN.
A responsibility regime that is based on effective control is desirable, but that does not mean that an injured party will necessarily have access to a remedy against the actor whom it deems responsible. The DARIO, like the Articles on State Responsibility for that matter, are concerned only with rules of responsibility, and remain silent on accountability mechanisms.Footnote 64 The availability of remedies for injuries suffered by acts or omissions of UN peace-support personnel may be particularly problematic, especially if the UN's position of quasi-automatic attribution of wrongful acts to the UN itself is followed. The UN is ordinarily considered to be immune from suit before the courts of its member states (on the basis of Article 105 of the UN Charter), while there is no internal UN judicial mechanism with proper jurisdiction to hear complaints lodged against the UN. Even where the responsibility of the contributing state may be engaged, typically by the courts of that very state, national troops may possibly enjoy immunity from claims relating to combat action,Footnote 65 or a state's human rights obligations may not apply extraterritorially.Footnote 66 This may potentially lead to the extremely undesirable situation that recognised human rights standards do not seem to apply to victims of violations committed in the course of UN-mandated peace operations, and that victims thus find themselves in a legal limbo.Footnote 67
This legal limbo can only be remedied if responsible actors – the UN and troop-contributing states – provide adequate mechanisms whereby victims can bring their cases. Both the UN and troop-contributing states may want to set up dispute settlement mechanisms that are endowed with more due process assurances than internal review boards. The UN, for one, could finally act on Section 29 of the 1946 UN Convention on Privileges and Immunities (‘1946 UN Convention’) – that the UN ‘shall make provisions for appropriate modes of settlement’ for disputes to which the UN or a UN official is a party. The establishment of a UN Standing Claims Commission, a quasi-judicial body, was envisaged in this section, but it never materialised. Instead, UN local claims review boards have been established in the conflict theatre itself, but these boards can hardly be considered to be judicial institutions.Footnote 68 Alternatively, such mechanisms can be made available outside the conflict theatre: a special mechanism may be established at the UN headquarters or claims may be brought in the domestic courts of the troop-contributing state (compare the claim brought by the relatives of some Srebrenica victims in the District Court in The Hague). As far as domestic courts are concerned, claims against the UN as an organisation can, of course, only be successful provided that the said courts do not automatically uphold the immunity of the organisation. It is suggested in this respect that domestic courts tie the grant of immunity to the UN to the availability of adequate dispute settlement mechanisms within the UN system (compare Section 29 of the 1946 UN Convention), in line with the ECtHR's case law in this regard.Footnote 69 The UN may enjoy immunity on the basis of Article 105 of the UN Charter and the 1946 Convention, but such immunity should be reviewed in light of the individual right to a remedy (Article 6 of the European Convention on Human Rights (ECHR), Article 14 of the International Covenant on Civil and Political Rights): only if the UN offers reasonably alternative means of dispute settlement should immunity be granted. To its credit, the Dutch Court of Appeal hearing the Srebrenica case against the UN did realise that the UN's immunity had to be reconciled with the right of access to a court, but unfortunately the alternative means of dispute settlement suggested could hardly be considered as adequate.Footnote 70A domestic avenue is only a second-best option but at least the risk of domestic litigation may exert pressure on the UN to establish adequate dispute settlement mechanisms where victims could file claims with a reasonable chance of success.
4. Effective Control in Srebrenica, Al Jedda, and Mukeshimana
In the previous sections, we have defended the standard of ‘effective control’ as the appropriate standard to allocate responsibility for wrongful acts committed in the course of UN peace operations. We have criticised the UN's position on the attribution of responsibility – and The Hague District Court's decision which was informed by it – since it fails to take operational realities into account and thus fails to serve a preventative function. We have also criticised the ECtHR's decision in Behrami, since the ‘ultimate authority and control’ standard it embraces similarly fails to do justice to effective operational realities. In Behrami, unified ‘effective’ command and control was not exercised by the UN – to which the Court attributed all wrongful acts that may have been committed by troops in Kosovo – but by NATO.Footnote 71 The ECtHR's observation that the UN Security Council had only ‘delegated to NATO … the power to establish, as well as the operational command of, the international presence, KFOR’Footnote 72 and thus that the UN ‘was to retain ultimate authority and control over the security mission’Footnote 73 is not relevant in this respect, as this observation relates to the institutional legality of the delegation rather than to the locus of responsibility for violations committed in the exercise of delegated powers.Footnote 74 What was relevant to apportioning responsibility was the level of control exercised by the national contingents. It was NATO and its member states which exercised effective control, and not the UN. This criticism, which was voiced in the literature in the immediate aftermath of Behrami,Footnote 75 has recently also been voiced by the UN itself, which suggested in its 2011 comments to the ILC that the rules of state responsibility should have applied to a Behrami-style UN-authorised, as opposed to a UN, operation.Footnote 76
Our argument in this section is that the said criticism of Behrami is reflected in post-Behrami court practice. Domestic courts in three different states – the Netherlands (Srebrenica, 2011), the United Kingdom (Al Jedda, 2007), Belgium (Mukeshimana, 2010) – have recently heard cases regarding the allocation of responsibility in UN peace-support (UN-authorised or UN-supervised) operations in Iraq, Bosnia Herzegovina, and Rwanda, respectively. These courts applied a proper ‘effective control’ standard, in line with Article 7 DARIO, and – finding that in all cases the state exercised effective control over the troops deployed in various conflict theatres – located responsibility with the state rather than the UN. The UK House of Lords’ decision in Al Jedda was recently confirmed by the ECtHR, the very court that rendered the misguided Behrami decision.Footnote 77 One is indeed tempted to identify a tendency in recent domestic courts’ – and thus state – practice in favour of the application of an ‘effective control’ standard, and in favour of the attribution of conduct to troop-contributing states rather than to the UN in peace-support operations in cases of effective control by the state.
4.1 Srebrenica
The Srebrenica case before the Dutch courts concerned the responsibility in tort of the Dutch government for the conduct of Dutch UN troops (‘Dutchbat’, part of the UNPROFOR mission) deployed in Bosnia in 1995, not for their failure to prevent the massacre in Srebrenica, but rather for their refusal to allow the family of a Bosnian interpreter and a Bosnian electrician working for Dutchbat to remain on the Dutchbat compound and to be evacuated together with Dutch troops – as a result of which they were killed by Bosnian Serb troops and paramilitaries commanded by General Mladic.
It is recalled that, at first instance, The Hague District Court dismissed the complaint on the grounds that the operational command and control of Dutchbat lay with the United Nations, rather than the Netherlands, and that Dutchbat did not back out of the UN command structure.Footnote 78 As a result, the responsibility of the Dutch government could not be engaged. This reasoning was open to criticism as, according to the facts of the case, in the period before the Srebrenica massacre, the Netherlands arguably retained a significant amount of control and command over the Dutch UN troops deployed at the compound close to Srebrenica. Moreover, to consider the national troops’ backing out of the UN command structure as the sole exception to UN responsibility, as the Dutch court maintained,Footnote 79 seems to have no basis in international law.Footnote 80 Again, effective control appeared to be the relevant standard, much more persuasive than legal minutiae of organisational hierarchy: in fact, since in peace operations national troop contingents are never fully part of the UN command structure, and the command may have been the responsibility of the troop-contributing state in the first place, the military themselves cannot really back out of the UN command structure. Of course, to the extent that, arguendo, they were indeed initially under the effective control and command of the UN, then the national troops’ backing out of the UN command structure through receipt of instructions exclusively from the Dutch government evidences that effective control was in fact being exercised by the Dutch.
In spite of the misgivings one can have about the District Court's decision, it nonetheless came as a surprise that, on appeal in 2011, The Hague Court of Appeal overruled the decision; after all, the latter decision reflected the UN's position, and any other decision might have serious financial consequences for the Dutch government.Footnote 81 Applying the effective control standard laid down in Article 7 DARIO,Footnote 82 it found that, given the circumstances of the case, Dutchbat's acts in respect of the Bosnian interpreter's family and the electrician were attributable to the Dutch government. The Court was of the view that after the fall of Srebrenica on 11 July 1995, both the UN and the Dutch government had control over Dutchbat's remaining task to assist and evacuate the Bosnian refugees, and the preparation of the complete withdrawal of Dutchbat from Bosnia. This was so for various reasons: (i) the Dutch chief of staff of UNPROFOR headquarters served in a dual role as a representative of the UN and the Dutch government; (ii) two high-ranking Dutch military men, together with the French UN Force Commander, had taken the decision to evacuate Dutchbat and the refugees; and (iii) the Dutch Minister of Defence had given specific instructions to Dutchbat. Since the Dutch government was so closely involved in the (preparation of the) evacuation, according to the Court, it had the power to prevent Dutchbat from ordering the Bosnian men to leave the Dutchbat compound – an order which, in the Court's view, also violated UN instructions to protect refugees as much as possible.Footnote 83 Because the Dutch government allowed the men to leave the compound and failed to have them taken to a safe area – as a result of which they subsequently met their death at the hands of Mladic's men – the Court concluded that the government had acted wrongfully as far as these men were concerned.Footnote 84
The Court of Appeal's decision in Srebrenica is based on an arguably correct understanding and application of the ‘effective control’ standard, and the preventative rationale which this standard serves. The Court inquired in detail into the factual circumstances that were present at the time, and located responsibility, at least in part, at the level of the Dutch government, given its involvement in the peace operation in Srebrenica, and the power to prevent the evacuation of the Bosnian employees. The link between the power to prevent and effective control was not explicitly made by the ILC.Footnote 85 But, as was also argued by Dannenbaum,Footnote 86 an actor's omission or failure to prevent does engage that actor's responsibility if, all things considered, it could have exercised control over the situation.
It could be gleaned from various passages in the judgment that, in the Court's view, not only the Netherlands but also the UN was to blame, possibly legally (under the effective control standard).Footnote 87 Both the Netherlands and the UN would then share international responsibilityFootnote 88 – which, as argued above, may reflect the command and control structure in typical UN operations. The likelihood of the UN's legal responsibility was not pursued by the Court, however, because the UN was considered to enjoy immunity.Footnote 89As regards the more general failure of the UN with regard to the fate of all Srebrenica refugees (beyond the specific individuals mentioned in Nuhanovic and Mustafic, who, as employees of Dutchbat, had a particularly close connection with the Dutch government), one may wonder whether the UN had committed sufficient resources to the UN mission in terms of air support, etc. The possibility that the Dutch troops might have required such reinforcement, and that had more UN support been forthcoming the Srebrenica massacre could have been prevented, represent facts which are material to the allocation of responsibility. Thus, it may be submitted that, alongside the Netherlands, it is reasonable to expect that the UN also exercised control over the situation in Srebrenica, and for that reason its responsibility should be engaged.Footnote 90
4.2 Al Jedda
In the immediate aftermath of Behrami, in 2007, the United Kingdom House of Lords decided the Al Jedda case, which concerned liability for violations committed against a detainee in Iraq (holding nationality of both the UK and Iraq) by UK troops who allegedly acted under UN auspices.Footnote 91 As, on the facts, Al Jedda's claim was denied,Footnote 92 he applied to the ECtHR, which decided in July 2011 that the applicant's right to liberty under Article 5(1) of the ECHR had been violated.Footnote 93
The facts of Al Jedda differed considerably from the facts of Behrami. UN Security Council resolutions may have governed the presence of international troops (Resolutions 1511 and 1546, in particular), but this presence was clearly not a typical UN peace-support mission of which the establishment was initiated by the UN, nor were the troops deployed to support an international territorial administration. One may wonder, however, whether those factual differences, emphasised by both the House of LordsFootnote 94 and the ECtHR,Footnote 95 should really matter from a conceptual point of view: the relevant criterion, after all, is the exercise of effective control, not the mode of establishment of a UN peace operation. Command and control may not always be exercised in the same manner in all operations in which the UN is involved, but it is the same standard of responsibility which governs the legality of acts committed during all UN peace operations: who exercises effective control over the conduct giving rise to the violation?Footnote 96
In spite of the UN framework within which the UK was operating – and which, in fact, allowed the UK to detain prisoners for security reasons – Lord Bingham, after analysing the factual background of the case, observed not surprisingly: ‘It cannot realistically be said that US and UK forces were under the effective command and control of the UN, or that UK forces were under such command and control when they detained the appellant.’Footnote 97 The ILC stated in its commentary to Article 7 DARIO that ‘[t]his conclusion [which located responsibility with the UK] appears to be in line with the way in which the criterion of effective control was intended’.Footnote 98
The standard of responsibility on which the House of Lords’ decision was based was subsequently endorsed by the ECtHR in Al Jedda, which also cited Article 7 DARIO in support.Footnote 99 But it is conspicuous that the ECtHR was visibly at pains to do justice to both the effective control standard of Article 7 and the Behrami ultimate control standard: it held literally that ‘the United Nations Security Council had neither effective control nor ultimate authority and control over the acts and omissions of troops within the Multi-National Force and that the applicant's detention was not, therefore, attributable to the United Nations’.Footnote 100 The reference to ‘ultimate authority and control’ is undesirable, as explained above, but it is understandable that the Court was reluctant to set aside, given its previous – and, after all, very recent – case law. Instead, it simply opted to ‘add’ the effective control standard as a review standard. The coexistence of two control standards may be confusing, but it is a step forward in that the Court seems at least to have understood the true meaning of Article 7 DARIO (without, however, going as far as to refer Behrami to the dustbin of history).Footnote 101
The semantic question remains, however, whether the abstract standards of ‘ultimate’ and ‘effective’ control really capture different situations. The national force commander of a peacekeeping or peace-enforcement contingent may have effective control over its troops, but lack the power to make any major strategic decisions if the UN commander is endowed with that power. In that case, the UN may be said to exercise ultimate and effective control. Ultimately, the locus of responsibility – and the control standard used to identify that locus – should be where decisions affecting the enjoyment of rights are taken. The ECtHR in Behrami may then perhaps not have adopted the wrong control standard, but simply wrongly applied it.
Because the UK exercised effective and/or ultimate control over its troops in Iraq, it was the liability of the UK rather than that of the UN that was engaged.Footnote 102 Therefore, the fact that UK troops acted under the auspices of the UN in Iraq in no way removed the UK troops’ own obligations under international human rights law vis-à-vis individuals over whom they exercised effective control. The question then arose, however, whether the UK may not have been under an obligation to intern Al Jedda pursuant to UNSC Resolution 1546.Footnote 103 This issue – which concerns the relationship between UN Security Council action and human rights obligations – will not be further dealt with here, as it is a primary rule question rather than a secondary one with which this article is concerned.
4.3 Mukeshimana
A third decision that correctly applied the effective control standard is the Belgian Mukeshimana interim decision (2010) rendered by a first instance civil court in Brussels.Footnote 104 This decision, regarding the apportioning of responsibility between the UN and Belgium for the failure of Belgian UN peacekeepers to prevent a massacre in Rwanda in 1994, emphasises the operational character of effective control and, unlike the Dutch Srebrenica decision, admits that in reality, in the midst of a UN peace operation (in this case MINUAR, the UN Mission for Assistance in Rwanda), as a result of the chaotic conditions in the field and corresponding communication problems, UN effective command and control over national troop contingents may shift to the troop-contributing state. In that case, the state is to be held responsible if wrongful acts were committed.
Mukeshimana is the first – interim – judgment in a tort case brought before a Belgian court against the Belgian state (and a number of Belgian commanders) by victims of the massacre following their evacuation from the compound where they had sought refuge. The court accepted the plaintiffs’ argument that control over the troops stationed at the compound no longer rested with MINUAR but was placed under the exclusive responsibility of the Belgian state. Examining the operational circumstances surrounding the peacekeeping operation in great detail, the court made it clear that it is immaterial whether command and authority over the UN peace operation may initially have been understood to rest with the UN commander. What is relevant is who, in the midst of the events unfolding, was actually exercising authority over the troops. In Rwanda, the MINUAR troops based at the compound were under the authority of the national force commander who consulted the Belgian army's chief of staff and not the UN commander.Footnote 105 Accordingly, responsibility for any internationally wrongful acts committed by the Belgian MINUAR troops lay – and should indeed lie – with Belgium as the troop-contributing state, and/or the Belgian commanders.Footnote 106
Precisely because the court inquired into who in fact exercised effective control and command over the troops, regardless of the actor's formal authority in the circumstances, its decision deserves considerable support in light of the control standard enunciated in Article 7 DARIO. It is unclear, however, whether this interim decision will be upheld at a later stage of the proceedings or on appeal.
5. Conclusion
The 2009 ILC Commentary to current Article 7 DARIO contained only a few references to state, institutional, and court practice that supported the ‘effective control’ standard laid down in that provision. In fact, most references undermined the standard. First, UN practice was cited at length,Footnote 107 but as argued above, the UN takes the view that, in principle, it always has effective control over UN troops, a view which ultimately renders Article 7 redundant. Second, the Behrami case, cited by the ILC,Footnote 108 was based on a standard of ‘ultimate control and authority’ rather than ‘effective operational control’. And third, the Dutch District Court in Srebrenica, also cited by the ILC,Footnote 109 failed to inquire in any in-depth manner as to who actually exercised control over the Dutchbat troops, assuming, just like the UN, that all acts and omissions of UN troops should be attributed strictly, as a matter of principle, to the United Nations. In fact, only the case of Al Jedda could be cited in support of Article 7 DARIO,Footnote 110 but that case was so specific – it concerned the responsibility of a state for acts carried out within the framework of a multinational military operation (Iraq) that was initially not authorised by the UN, and with which the UN later entertained tenuous links only – that it could hardly serve as a precedent.
Ever since, however, practice supporting the rule in Article 7 DARIO has been growing. In July 2011, the European Court of Human Rights confirmed the UK House of Lords’ Al Jedda decision, and juxtaposed the ILC's ‘effective control’ standard to the Behrami ultimate control standard. At about the same time, The Hague Court of Appeal overruled the District Court's decision in Srebrenica; applying the ILC's ‘effective control’ standard, it held that Dutchbat's acts were attributable to the Netherlands rather than to the UN only, given the intensity of Dutch government involvement in Dutchbat. And in December 2010, a Belgian first instance court provisionally ruled that the acts of the Belgian MINUAR troops, deployed in Rwanda in 1994, were to be attributed to Belgium rather than (only) the UN, given the effective control exercised over those troops by commanders of the Belgian army rather than UN commanders.
Clearly, there is a tendency among the judiciary to apply the standard of ‘effective control’ as the applicable yardstick to apportion responsibility between the UN and its member states. This process may have been set in motion by the ILC's (current) Article 7 DARIO, which may in due course start to reflect customary international law.Footnote 111 From a policy perspective, this process is highly desirable, as it locates responsibility with the actor who is in a position to prevent the violation.