3.1 Introduction

On board of a Finnair flight, in October 2021, the seat pocket contained Finnair’s flight menu, advertising products available for on-board purchase. The products on offer were standard enough: coffee and tea, sodas, snacks, and beer and wine. The (exorbitant) price level too was not very surprising. But what was surprising was a small-print disclaimer: having first stated that prices and selection may be subject to change, it continued: ‘Finnair is not responsible for misprints.’Footnote ¹ On some level, this is understandable: the printing has probably been outsourced to a sub-contractor, or perhaps even to a sub-contractor of the sub-contractor, or a further sub-contractor thereof. At some point it becomes difficult to keep track, even for the original assignor. On the other hand: the flight is a Finnair flight; the menu is offered by Finnair, and the goods are purchased from Finnair flight attendants – why shouldn’t the proverbial buck stop with Finnair? And if not with Finnair, then with whom? If there were a misprint, to whom should the passenger complain?

What applies to many business settings these days, characterized by the involvement of multiple actors in global supply and value chains and joint ventures, also applies to politics generally, and therewith to international organizations and their activities – including an organization such as the International Organization for Migration (IOM). Often enough, international organizations are involved in projects together with a multitude of other actors, some closely related to them (their member states, for instance), others more distant, from other international organizationsFootnote ² and funders to co-financiers to service-providers.Footnote ³ And this cannot but affect the topic of the accountability of international organizations, all the more so as shifting responsibility onto others is a useful strategic device. In what follows, I will first set out why international organizations law has difficulties handling accountability, delving a little into the history (Section 3.2) and epistemology of international organizations law (Section 3.3). Sections 3.4 and 3.5 take a more in-depth look at the most authoritative accountability regime, the ARIO, developed by the International Law Commission; succeeded by a closer look at the mechanisms available at IOM (Section 3.6). Section 3.7 concludes.

The argument I will make is a general argument, equally applicable (mutatis mutandis) to IOM as to the World Bank, or the World Health Organization or even the European University Institute. While it is arguable that IOM has no strong human rights protection or humanitarian mandate, this circumstance alone is unlikely to affect its legal accountability – the problems with accountability of international organizations under international law go much, much deeper. And by legal accountability (not quite a term of art perhaps), I mean something like utilizing a (more or less) legal mechanism to test the acts of an international organization against (more or less) legal standards. This may be done before a court, but may also involve internal accountability mechanisms. Those standards, in turn, do not simply comprise the entire corpus of international law, but are limited, it is generally agreed, to the treaties that international organizations are parties to, to international legal rules that have become internalized, and to the ‘general rules of international law’.Footnote ⁴ There is consensus that this is an authoritative enumeration, but precious little agreement on what this entails (it will be further discussed below).

3.2 The Vacuum Assumption

The accountability of international organizations under international law has proved a difficult topic, albeit of relatively recent provenance. For more than a century, from the 1860s to the 1980s, the topic did not exist. International organizations were supposed merely to interact with their member states: legally as well as theoretically, a vacuum was drawn around the relationship between international organizations and their member states, and the idea of holding international organizations to account simply never came up, at least not with respect to third parties. After all, since international organizations were not supposed to deal with third parties, issues of accountability towards third parties could not logically arise – quod erat demonstrandum. While some organizations were created to take care of individuals, those individuals were conceptualized as merely the objects of organizational activity – not as interlocutors or partners in any meaningful way.

There was only one exception, and it was not immediately related to third parties: member states could control their organization, if only they could muster the unity to suggest that the organization had overstepped its powers, acted ultra vires, or maybe violated some internal rule or other. This way of thinking was behind the 1962 Certain Expenses opinion of the International Court of Justice, with France and the USSR contesting the legality of peacekeeping ‘recommended’ by the General Assembly (GA) of the UN. This, they claimed, effectively meant the GA had been acting ultra vires, and how could states be expected to help finance ultra vires activities? The ICJ disagreed, but without taking a firm principled stand: activities ultra vires the GA could still be intra vires the UN at large, and thus could be viewed as legitimate expenses, to be provided for under the regular UN budget. Whether peacekeeping was ultra vires the UN itself was a question not further addressed,Footnote ⁵ and the idea that the GA could sponsor peacekeeping was in line, the Court suggested, with the idea that the UN Charter merely assigned ‘primary responsibility’ for peace and security to the Security Council. And this made it possible to suggest that the GA exercised a secondary responsibility.Footnote ⁶

So, the member states can hypothetically control the acts of their international organizations: if the members together disapprove of an action or a policy, then the organization can be compelled to mend its ways. There are two obvious drawbacks though. The first is that for this to work, the members must all sing from the same hymn sheet: if only one or two think the organization does wrong, then control will be out of reach. What then typically happens is that individual member states take the law in their own hands and try to exercise political pressure. This may take place by withholding their contributions (a weapon all the more potent when the organization is hugely dependent on a single member state, as with IOM vis-à-vis the United StatesFootnote ⁷); by threatening to withdraw from the organizationFootnote ⁸ or even by ousting the director-general.Footnote ⁹ And then there are other pesky ways to make life difficult: delaying visa applications for organization staff, not allowing aircraft to land or not allowing staff or management into the country, that sort of thing.

The second drawback is that this form of control still assumes the vacuum drawn around the organization and its member states: it is of little use to third parties in terms of their ability to demonstrate or advance their own accountability claims. An international organization breaching a treaty commitment towards a third party, or a commercial agreement with a service provider, will not, given the assumed vacuum, incur accountability. And even more seriously, when the organization commits a wrong to an individual, it has historically proven difficult to address the matter, let alone to find redress. This is partly a matter of immunities law (international organizations can typically invoke immunity for their official acts, and are not afraid to do so), but it goes deeper: in a setting where there exist no third parties, with a legal system which cannot think about third parties, accountability towards third parties will remain elusive.Footnote ¹⁰

Against this background, it is no coincidence that the first academic attempts to come to terms with the accountability of international organizations remained unsuccessful. Attempts in the 1950s by EagletonFootnote ¹¹ and by GintherFootnote ¹² in the 1960s came to naught (although Ginther coined the glorious term Durchgriffshaftung – literally, something like ‘see through responsibility’ – to discuss the responsibility of member states for acts of the organizationFootnote ¹³), and quickly moved to the possible responsibility of member states for acts of their organizations. For while practically speaking, international organizations can and do affect third parties, the law had no way of handling this, so the idea that international organizations could be accountable in their own right, in their own name, as independent actors with their own international legal personality, just did not arise. And it could not even arise: in a rather literal sense, the thought had not yet been thought.Footnote ¹⁴

That this situation was problematic became clear with the International Tin Council (ITC) litigation in the mid-1980s. The ITC, an international organization based in London, became insolvent; banks and others claimed their money back; the ITC was unable to make good on its loans, and as a result several creditors started proceedings against the ITC’s member states. This however, was unsuccessful before the UK courts (where the litigation played out): if international organizations are separate persons, it follows that their accountability is separate from that of their member states. Accordingly, member states cannot be held liable for the acts of their international organizations. The ITC litigation made waves: the legal discipline started to realize that international organizations could actually do wrong in their own name – in this case, defaulting on debts – and perhaps it is no coincidence that the wake-up call related to large sums of money rather than the suffering of ordinary people. And there was nothing the law could do about it – or was there?

Various pens were moved, first of all to confirm the position that member states are and should be shielded.Footnote ¹⁵ Others went a bit further and started to explore arguments of principleFootnote ¹⁶ and, more inductively, the relevant case law of international and domestic tribunals.Footnote ¹⁷

Others started to search for administrative precepts which could possibly be applied to instances of global governance, including the acts of international organizations. Most prominent among these is the Global Administrative Law approach (GAL), tapping into administrative law thinking in the hope of finding ideas that could be used in the ‘global administrative space’: this would include such ideas as participation in decision-making, providing reasons in judicial judgments, and using proportionality.Footnote ¹⁸ Still, this did not solve all issues, partly because in order to hold international organizations to account, there must be standards according to which they can be held to account. Borrowing administrative principles from European and US traditions was considered a bit too Western-centric, and even within Europe there are fundamental differences about the role and function of public law: some view public law largely as a check on overzealous governance while others view it rather as enabling governance.Footnote ¹⁹ Moreover, the GAL approach remained unable to resolve one of the fundamental issues: why, unlike states, are international organizations bound to respect rules they have not consented to?

Even the ILC, never the most agile body, stepped in, and between 2001 and 2010 developed a regime on the international legal responsibility of international organizations, the Articles on the Responsibility of International Organizations (ARIO). And the ILC put its finger on the sore spot. It suggested, sensibly enough, that organizations should be held responsible for their internationally wrongful acts, and these are thought to consist of two elements: a violation of an international legal obligation incumbent on the organization that is attributable to the organization. Both elements prove to be extremely difficult. This raises a further question, to be discussed in Section 3.3: why is international organizations’ accountability so difficult?

The very term ‘accountability’ (and related terms like ‘responsibility’ or ‘liability’) already carries a strong suggestion that the entity concerned has done something questionable. At issue is the control of the acts of the organization, but whereas ‘control’ is a relatively neutral, unloaded term that at most suggests that the organization needs someone in charge, accountability and related terms are considerably more politicized. Put differently, the very term ‘accountability’ presupposes what often needs to be proven: that international organizations do wrong – ‘control’, by contrast, raises the possibility of wrongdoing, but without having reached that conclusion just yet.

Furthermore, accountability (and related terms) is usually backwards-looking: it makes sense to speak of controlling what an international organization plans to do tomorrow, but it makes less sense to speak of holding it accountable for what it plans to do tomorrow. Linguistically, it would seem odd to incur accountability for something that has not yet taken place, although in pledging to respect particular principles, such as human rights and humanitarian standards, international organizations create expectations regarding their future behaviour, and may be called to account for deviations from these commitments. The point for present purposes is not that the term accountability is out of place – it is merely to suggest that the term itself is based on certain assumptions which may or may not withstand further scrutiny.

Relatedly, the question arises of what and whose standards are considered of relevance. The ILC focuses on international legal obligation, and that is fine as far as things go. But different constituencies might rely on different and possibly contradictory standards of accountability, reflecting their own policy preferences. Put concretely, donors to specific IOM projects may rely on different standards than migrants do, whose priorities may also differ from those of the member states collectively and from those of (often foreign-based) civil society organizations.Footnote ²⁰ This is likely to result in confusion and a leaking away of accountability – how to decide whose preferences weigh heavier?

Relatedly, it makes sense to think that organizations should be held responsible for misconduct, but often the problem lies elsewhere: it is often claimed that the organization should be held accountable for acts done in the course of doing its job. This comes in broadly two variations. First, in the exercise of a task, the organization can stumble on other, external, standards, to which it may or may not be bound as a matter of law. The classic example is the lengthy discussion about the World Bank and human rights, with the Bank caught between its own constituent instrument and a number of other possible standards supported by different stakeholders. This applies also to international organizations which have publicly stated to respect human rights, as IOM has done, unless one could claim that the human rights at issue are peremptory norms of international law (jus cogens). This may apply to some human rights norms (the torture prohibition, e.g. or the non-refoulement rule), but is unlikely to apply to most human rights norms. The net result then is a conflict of norms, and those often defy easy solutions, even if formerly external standards are transformed into ‘internal rules’. They will, often enough, need to be balanced against other international rules.

The second scenario is where the organization causes damage (or contributes to it), without violating any particular international legal obligation. Here a standard scenario is that of the UN inadvertently bringing cholera to Haiti. The UN may have made some debatable decisions, such as contracting a local waste management company, likely for cost reasons. And most assuredly the UN should have issued an apology for a catastrophe happening on its watch. Still, it seems to have followed its own procedures for preventing the spread of communicable disease which had been working quite well for half a century, with a three-month window between testing and deployment. At worst (and not very plausibly, given the existence of these procedures), the UN can be accused of negligence, but how to give this hands and feet in international law? To some extent, this gets done by invoking an obligation to exercise due diligence, but due diligence itself remains rather elusive contents-wise, and it often remains unspecified why, as a matter of positive law, international organizations would be under an obligation to exercise due diligence.Footnote ²¹

3.3 Tropes Underlying the Law

As noted, international organizations were imagined as entities without external relations, let alone legal interactions with third parties. Whether this was ever tenable is beside the point (and really, it never was tenable), but what is relevant is that this became a very strong assumption – where actors were not expected to interact with the outside world, the legal system need not make arrangements for this; and by the time external engagement became topical, the vacuum assumption was firmly in place.

Behind the accountability discussion lie deeper tropes. If there is a tension between external standards and the mandate of an international organization, why not simply settle this in favour of external standards? After all, that is what happens with states: states cannot use their internal set-up as an excuse for violating international law. So why are things different with respect to international organizations?

Here the topoi underlying international organizations law make an appearance.Footnote ²² When international organizations are discussed, the adjective gets emphasized: international organizations are viewed as manifestations of the ‘international’, rather than as a particular kind of ‘organization’. Doing so taps into a number of related tropes. First, for many (and especially international lawyers), the ‘international’ has a specific attraction. The ‘international’ is somehow regarded as superior to ‘parochial’, internationalism is considered better than nationalism. International lawyers are not alone in this: the thought can be traced back centuries, to KantFootnote ²³ and many writing before him. Few might opt for world government, but somehow internationalism is synonymous to peace, to harmony, to universal understanding.

This in turn borrows from a deeper idea: cooperation is considered superior to the absence of cooperation. Whether the proposition is generally tenable, is debatable (torture too depends on many people working togetherFootnote ²⁴), but for that no less forceful. Without cooperation, life would be ‘nasty, brutish and short’. The topos is a strong one, deeply engrained and embedded in political thought. And that entails that for international lawyers, a soft agreement is always preferable to no agreement at all: voilà the most obvious explanation for the popularity of ‘soft law’.Footnote ²⁵ And international organizations, as manifestations of international cooperation, can accordingly do little wrong – almost literally.

With respect to international organizations, there is a further topos to consider: the idea that ‘the end justifies the means’. This applies with particular force to international organizations; these, after all, are almost by definition set up to reach a certain end. The very core of international organizations law is that they exercise a function, a task, set to them by their member states. This even applies to organizations whose goal is very abstract and somewhat contested: think of the European Union’s goal of becoming ‘an ever closer union’. This is impossible to demarcate with any precision, but important it is nonetheless considered to be.

It is no coincidence that Jellinek, writing in 1882, used the term Verwaltungszweck to discuss international organizations, with the word Zweck translating as goal, or end. International organizations have an end (as organizations generally cannot work without a goal or telosFootnote ²⁶), and whatever contributes to that end should be given pride of place, while whatever might obstruct the achievement of the end should be brushed aside. Previous generations have intuitively recognized this, and have used biblical imagery to underscore the point. Claude gave his highly popular post-war textbook on international organizations the title Swords into Plowshares, in one linguistic stroke summing up the idea that peaceful order can be born out of the anarchical international society if only we let international organizations do their job.Footnote ²⁷ And Singh, a future President of the International Court of Justice, even went one better, attributing to international organizations generally a crucial role in the ‘salvation of mankind’.Footnote ²⁸ The message is clear: let international organizations do what they were set up to do, and the world will be a better place.Footnote ²⁹ The idea holds a strong place in the collective minds of specialist lawyers: international organizations should act without impediments because they will bring us the salvation of mankind – and who would possibly want to stand in the way? This is irrespective of the precise international organization concerned: the oil cartel that is OPEC or the military alliance of NATO benefit from the positive image of international organizations founded on the above-mentioned tropes, as does IOM. The law has been unable to differentiate between organizations under reference to their perceived public ethos, and the precise constitutional mandates do not alter the picture. The topoi operate at a far deeper level, and even a hypothetical nasty international organization would be considered to manifest cooperation and represent ‘the international’, although in the case of an obviously nasty organization one might pause at thinking that the end would justify the means.

If the above is accurate, then it is no wonder that the law has problems thinking of international organizations as being accountable to third parties: the end, after all, justifies the means, and the end is considered so important that a little collateral damage is considered perfectly acceptable. On this line of thought, if UNHCR runs a refugee camp and decides to withhold food from those who seem a bit obstinate, that is considered quite acceptable: the obstinate interfere with the functioning of UNHCR.Footnote ³⁰ And if the World Bank ends up displacing thousands of people in the name of a development project, again, the end justifies the means. Most of these topoi have a natural counterpart (local over global; sometimes cooperation is bad; some means are intrinsically bad), but the point for present purposes is precisely that these topoi strongly influenced – and still influence – the way international lawyers think about international organizations.

3.4 Internationally Wrongful Acts: Some Problems

But even without considering the above topoi, it will be difficult to hold international organizations to account. There are few institutional external arrangements available to enforce such obligations as international organizations may have. Typically, international organizations enjoy a large measure of immunity from legal proceedings before domestic courts. This applies also to IOM, which under Article 23 of its Constitution can claim a functional level of privileges and immunities. The text is somewhat ambivalent, with paragraph 3 of the same Article suggesting that the privileges and immunities ‘shall be defined’ in agreements between IOM and states. This can be seen as meaning that there are no privileges and immunities in the absence of further agreements, but this is difficult to reconcile with the wording of paragraph 1, stating that IOM ‘shall enjoy’ privileges and immunities to the extent necessary for its functioning – and this would seem not to require further action. That said, calling for further action is functionally expedient, in that positing the absence of privileges and immunities suggests IOM may sometimes be impeded in its work, and detailed agreements will contribute to legal certainty.Footnote ³¹

Moreover, international organizations cannot be made to appear before the International Court of Justice (ICJ), as only states can be parties to proceedings before the Court. And much the same applies to other international tribunals. There have been some arbitrations before the Permanent Court of Arbitration (PCA) involving international organizations, but the awards have invariably been kept confidential. Hence, it is difficult to get a sense of which rules were applied, how responsibility (if any) was conceptualized, et cetera.Footnote ³² This does little to boost confidence in closing the widely perceived remedies deficit.Footnote ³³ And sometimes, quasi-judicial panels are set up to address specific instances or episodes of governance, or limited aspects thereof: think of the Kosovo Human Rights Advisory Panel, set up in the aftermath of the UN exercising governmental tasks in Kosovo, or the EU’s Human Rights Review Panel, accompanying the EU’s exercise of governmental tasks in Kosovo.Footnote ³⁴ Still, these remain exceptions.

For the better part, the options available tend to be internal to international organizations: compliance mechanisms, ethics offices, departments of institutional integrity. Useful as these may be,Footnote ³⁵ they remain internal mechanisms, typically testing the activities of the organization concerned against internal standards. These standards may, but often do not, reflect international legal standards.Footnote ³⁶ But even if it were possible to identify available remedies, two problems of a more principled nature remain. The first of these pertains to the basis of obligation in international law when it comes to international organizations; the second concerns attribution.

The various attempts to formulate accountability standards for international organizations invariably have problems in coming to terms with the basis of obligation. The ILC’s ARIO specify that international organizations can only be held responsible in international law for their internationally wrongful acts, consisting of two elements: a violation of an international legal obligation incumbent on the organization, and attributable to the organization. And this raises two obvious questions: how do organizations incur international legal obligations, and when exactly are acts attributable to them? ARIO deal extensively with the latter question (more on this below), but not so much with the former; hence, guidance must be found elsewhere. In the WHO-Egypt advisory opinion, the ICJ held in 1980, somewhat in passing, that international organizations incur international legal obligations in three distinct ways:Footnote ³⁷ they are bound by the treaties they are parties to; by their internal rules (and these may reflect international law) and by what the Court termed, purposefully one may assume, the ‘general rules of international law’.Footnote ³⁸

International organizations conclude a variety of treaties. Nigh-on all international organizations will have concluded a headquarters agreement with their host state, and many will conclude operational agreements in their spheres of activity: troop-contributing agreements, mission agreements and status-of-forces agreements in the case of the UN; loan agreements in the case of the World Bank, et cetera. But participation of international organizations in multilateral treaties of a quasi-legislative nature is almost non-existent, and even more so if the EU (the only organization with a proper foreign policy, if it can still be considered an international organization to begin with) is excluded. International organizations are neither parties to human rights treaties, nor to humanitarian treaties or to environmental protection treaties.Footnote ³⁹ And what applies to international organizations generally, applies to IOM as well – it is not a party to any multilateral convention of the sort mentioned above.

It is not uncommon for international organizations to have internal instruments reflect international law. The World Bank will generally be mindful of human rights (as will other international organizations: very few of them commit torture, practice slave labour, or stifle freedom of religion), while the UN Secretary General in the 1990s issued a Bulletin declaring that the UN will apply the ‘fundamental principles and rules’ of international humanitarian law.Footnote ⁴⁰ Laudable as the latter may be, it nonetheless provides the UN with considerable wriggle room in concrete cases: it is not bound by the letter of the Geneva Conventions. Potentially important for present purposes, moreover, is that in 2013 the UN adopted a Human Rights Due Diligence Policy (amended in 2015) which, so it is argued, ought to be respected by entities related to the UN, including IOM.Footnote ⁴¹

The most controversial source listed in the WHO-Egypt opinion, however, is the reference to the ‘general rules of international law’. Many observers have taken this as a reference to ‘customary international law’,Footnote ⁴² but doing so is unpersuasive: had the Court wanted to refer to the entire corpus of custom, it could have done so explicitly. The better view is that the Court’s words refer to the ‘secondary rules’ of the legal system: those addressing the creation and application of primary rules.Footnote ⁴³ It would be difficult to imagine that international organizations could escape from general notions of treaty-making, or the general rules on jurisdiction; but it is also unlikely that, e.g. the International Civil Aviation Organization would be bound by the entire corpus of customary international law, regardless of whether it has in some way consented.Footnote ⁴⁴ The one possible exception is jus cogens (peremptory norms from which no derogation is permitted, such as the prohibition of genocide), but this follows from the very nature of jus cogens: it has to be binding on all actors, including international organizations; otherwise it cannot be considered jus cogens.

If the basis of obligation is difficult to capture, no less problematic is the idea of attribution. To put it bluntly: international organizations rarely have their own police officers, customs officers, and the like: they often depend for implementation of action on cooperation by their member states. Plus, in turn, their decisions are often traceable to some or all member states, and could (generally) not be taken without some member state involvement. At the very minimum then, international organizations can rarely act in full independence from member states. But there is more to it still: often enough, international organizations participate in projects in which others also participate. Well-known is the collaboration in the field between IOM and UNHCR, often also involving governments and other actors. In a development project, participants may include private banks, construction companies, local governments, and multilateral development banks. In other cases, such as peacekeeping, it may involve not just national troop contingents but also transportation companies, waste management providers, and yet other participants, including regional organizations. Hence, it is often difficult, perhaps impossible, meaningfully to distinguish between the various participants in attributing behaviour.

3.5 A Bird’s Eye View on ARIO: Answering a Different Issue

The ARIO are based on several assumptions about their practical effect. Above, it was already noted that they require a violation of an international legal obligation incumbent on the organization (rather than its member states), and this violation must be attributable to the organization concerned. Both elements, it was argued above, will rarely materialize, and they will even more rarely materialize at the same time.

Some scenarios seem obviously to engage the responsibility of the organization concerned. One can easily imagine, for instance, that pushbacks operations engaged in by Frontex, the EU’s border agency, will possibly engage the EU’s responsibility under international law. Pushbacks may in certain circumstances violate the prohibition of non-refoulement (often seen as an example of jus cogens,Footnote ⁴⁵ and therefore binding on the EUFootnote ⁴⁶), and Frontex is an agency of the EU – hence, responsibility is prima facie likely.Footnote ⁴⁷ Likewise, mistreatment of refugees by UNHCR staff or IOM staff running a refugee camp or similar settlements will prima facie engage the organization’s responsibility, as will sexual abuse by UN peacekeepers.

And yet, things are not entirely clear. One of the curiosities behind ARIO is that their application is premised on classical international legal thinking: Articles 43 to 49, regulating the possibilities for invoking ARIO, are limited to sketching the circumstances under which responsibility can be invoked by a state or an international organization. Systemically, this makes eminent sense: international organizations, by and large, only hold international legal obligations towards either states or other international organizations, so it stands to reason that these two categories of entities are the ones upon which the ARIO are premised. Put differently, under classic international law as it applies to international organizations, IOM has the capacity to conclude an agreement with, say, Uzbekistan and subsequently breach it; and IOM has the capacity to conclude an agreement with, for example UNHCR. And should customary international law apply to international organizations to begin with, it will be in the form of obligations owed towards other states and other international organizations.Footnote ⁴⁸ As a result, it is no surprise that ARIO discuss the circumstances in which responsibility can be invoked by directly injured states and international organizations; that it prescribes that states and organizations give notice when they invoke responsibility; that ARIO refer to the general admissibility criteria known to inter-state international law (nationality of claims, and exhaustion of local remedies); that states and organizations can lose their right to invoke responsibility; that it provides for invoked responsibility by a plurality of states or organizations; and that it eventually provides for responsibility to be invoked by states and organizations that are not directly injured. The underlying model is that of classical international, inter-state, law, which is limited to addressing claims between states, and to those cases where private complaints come to be owned by the state of nationality of the complainant. The only concession concerns the circumstance that under ARIO, international organizations too can be part of the system, and the only (minor) departure from the classic model consists of the possibility to invoke responsibility on behalf of the community interest.

But what has gone missing here is the circumstance that in the twenty-first century, the most problematic situations are not those where IOM violates a treaty obligation towards Uzbekistan or UNHCR, but where organizations exercise public power: where Frontex engages in pushbacks; where IOM runs a migration processing centre; where UNHCR staff decides on refugee status applications, where the UN exercises governance and policing powers.Footnote ⁴⁹ It is here that ARIO are found wanting, resting content with the savings clause of Article 50, suggesting that ARIO is ‘without prejudice’ to entitlements private or legal persons may have to invoke ARIO. Again, in systemic terms this makes sense, and yet, it also suggests that when most needed, ARIO retreat. Private persons with a grievance against an organization need to find another legal basis for invoking responsibility – the individual having been badly served by IOM needs to identify a different legal basis. This, in turn, is harmonious with Article 33 of ARIO, suggesting that rights ‘may accrue directly’ to individuals or legal persons. By way of example, the official ARIO Commentary mentions obligations of organizations arising out of employment, and the effects of peacekeepers’ breaches on individuals.Footnote ⁵⁰ But this ignores that while for the former there might be judicial mechanisms available in the form of administrative tribunals, this does not apply to the latter: confronted with allegations concerning the activities of peacekeepers, international organizations will be quick to invoke their immunity from suit.

If the first relevant assumption underlying ARIO is the classical inter-state model of international law, oblivious to the exercise of public authority by international organizations, the second is equally problematic, and harks back to the problem of attribution. The basic idea, understandable enough in a liberal society where actors are supposed to be autonomous and thus to be held responsible for their own actions, is that responsibility can always be carved up between those participating in a wrongful act. And in theory, or ex hypothesi perhaps, it can: one can make fine distinctions and yet finer distinctions about how actors collaborate and how this affects ‘their’ contribution to a wrongful act, and this is precisely what ARIO aim to do. It contains over a dozen Articles on attribution in one way or another or, put differently, around 20% of the ARIO is devoted to attribution. The least problematic of those are Articles 6 through 9, largely addressing the acts of international organizations themselves and suggesting that acts of an organization’s organs and agents are attributable to the organization.

Articles 14–19 see to divided responsibility: an international organization can incur responsibility for aiding and assisting another entity in committing a wrongful act – hence, it is not excluded that IOM would incur responsibility for training the Libyan Coast Guard and providing it with equipment and infrastructure.Footnote ⁵¹ Organizations may also incur responsibility for directing and controlling such an act; for coercing another entity in such an act; for using member states to circumvent obligations; and as members of another international organization. And Article 19 underlines that this is ‘without prejudice’ to the separate responsibility of other international organizations or states. The model, therewith, is one of ‘carved-up responsibility’: each and every act can presumably be broken down into smaller pieces; for some of these the organization will incur responsibility, for some others a collaborator will incur responsibility. The same presumption underpins a final set of articles, Articles 58–63, addressing the partial responsibility of states for the acts of international organizations and largely mirroring Articles 14–19: aiding and assisting by states; direction and control by states; coercion by states; circumvention of obligations resting upon states, and express or implied acceptance of responsibility by states, and again ‘without prejudice’ to the possible responsibility of the organization concerned or any other state or organization. The message then is clear: each and every single wrongful act can be broken down, divided, parcelled out. In a literal sense, there is no ‘sharing’ of responsibility envisaged, as each participant can potentially be held responsible for its own contribution. In other words: a scenario in which IOM helps to run a detention centre in Libya and is financed, in part, by the EU, would cause serious intellectual difficulties, for how to break this down into manageable bits of activities that might incur the responsibility of the various participants?Footnote ⁵²

Hence, the question arises: how realistic is it to think of parcelled responsibility? Its provenance is understandable: the philosophical basis of acceptable politics (and therewith law) is individualist and liberal, and has been for centuries.Footnote ⁵³ It is considered unfair (with minor exceptions) to punish A for acts of B or C, and thus there is a strong philosophical imperative to divide wrongful acts into a multitude of component parts for which a multitude of different actors can be held responsible. But in the real world, such clear-cut divisions are not always possible or plausible and, what is more, many have discovered that this liberalism invites them to artificially assign tasks to different entities, each with their own sphere of responsibility – this is how Finnair can claim, selling products on a Finnair flight and with prices listed on a Finnair menu, that responsibility rests elsewhere, for responsibility can always be made to rest elsewhere, either upwards (with the assignor) or downwards (with the sub-contractor).

3.6 An Excursion into IOM Mechanisms

Even though IOM is not legally bound to any human rights convention, the understanding is that at the very least, by concluding the 2016 IOM-UN Agreement, it bound itself to respect human rights, broadly speaking.Footnote ⁵⁴ The Agreement provides, after all, in Article 2, paragraph 5 that IOM ‘undertakes to conduct its activities’ in accordance with the purposes and principles of the UN and with ‘due regard to the policies of the UN furthering’ these purposes and principles, as well as ‘other relevant instruments in the international migration, refugee and human rights fields.’ What exactly this means in ordinary language is not entirely clear (and that is probably no coincidence), but at least it would seem to suggest that IOM has committed itself to act with a human rights sensibility.

In the virtual absence of external accountability mechanisms such as courts, IOM has developed some internal mechanisms to hold it to account, but it should be noted here that the term accountability in itself is versatile, and covers many forms of control.Footnote ⁵⁵ Thus, IOM has an Office of the Inspector-General, which can evaluate the acts of individual IOM officialsFootnote ⁵⁶ and is otherwise engaged in auditing IOM’s country offices or particular policies, but mostly in terms of effectiveness, understood in terms of whether the policies are effective in achieving their stated aims, or whether the country offices are run effectively from a bureaucratic perspective. And this has fairly little to do with how accountability is usually conceptualized in discourses surrounding international organizations. Similarly, like so many other international organizations, IOM as an employer has accepted the jurisdiction of the International Labour Organization Administrative Tribunal (ILOAT), which serves as the tribunal deciding staff disputes for some sixty international organizations. It has done so since 1999, and has thus far (late 2021) been involved in around fifty ILOAT cases which, given the circumstance that IOM employs some 15,000 people, is very decent.Footnote ⁵⁷ That said, a report ranking the internal justice systems of a number of international organizations is not very impressed: it ranks IOM 29th out of the 35 organizations scrutinized, which is all the more problematic, perhaps, as IOM is the fourth largest employer of the organizations covered.Footnote ⁵⁸ Either way, its activities as employer are not directly related to the more usual conception of accountability (focusing on the substantive activities of the organization rather than its role as employer), and neither is the work of IOM’s Office of the Ombudsperson.

The closest to accommodating regular accountability concerns at IOM is the Ethics and Conduct Office, providing counsel, promoting ethical awareness, reviewing allegations of retaliation and recommending protective measures. This too is not exactly a promise to act in conformity with generally accepted human rights standards, but comes somewhat closer. In order to give effect to this, the IOM website even offers a Confidential Reporting Form, offering individuals the chance to complain about fraud and corruption and misuse of resources (again, perhaps more useful to the organization than to the complainant) but also about harassment, retaliation and sexual exploitation and abuse.Footnote ⁵⁹

In 2020, IOM summarized and streamlined its accountability policies by means of a newly established Accountability to Affected Populations framework,Footnote ⁶⁰ realizing that its activities may have broader effects than merely on those who benefit from IOM’s work. While the document stipulates to be based on principles such as ‘do no harm’, non-discrimination, and zero tolerance for sexual abuse and exploitation, at no point does it claim that IOM will respect particular international legal instruments. The document comes closest in pledging that IOM’s crisis-related operations will adhere to ‘the humanitarian principles of humanity, impartiality, neutrality and independence in the delivery of its humanitarian response’.Footnote ⁶¹ When it comes to data protection, moreover, it adheres to its own data protection principles and those of the UN, rather than those promulgated for more general use, such as the EU’s General Data Protection Regulation.Footnote ⁶²

3.7 Conclusion

If and when IOM does wrong, it will be difficult to hold it to account under international law. This is partly because few external accountability mechanisms are available, but the problem (if that is what it is) runs much, much deeper than merely the absence of suitable mechanisms. It is written into the DNA of international organizations law that accountability will be difficult to achieve, regardless of the availability of mechanisms, the existence of privileges and immunities, and related matters. The heart of the matter is that the underlying framework does not allow for accountability, from a legal perspective at least: the ‘software’ of putative accountability schemes is structurally incompatible with the ‘hardware’ of functionalist legal theory and functionalist international organizations law. Whether other, perhaps less ‘legal’, mechanisms would fare better remains unclear. Recognized approaches propagated in the public administration literature (thinking of IOM as having ‘clients’ that it would be accountable to in accordance with a market model, for example, or enhancing possibilities for participation by stakeholders in decision-making and implementationFootnote ⁶³) do not appear to be very practical when it comes to international organizations generally, and much less so in times of urgency and crisis.

The only possible way out is not to have more rules; is not to have more tribunals; is not the lifting of immunity; the only way out, instead, is to re-think international organizations law from the ground up. Neither changing the IOM Constitution nor creating more internal mechanisms will do the trick as long as the ‘operating system’ is not capable of accommodating accountability towards third parties. As long as the law is dominated by the vacuum assumption, established over a century ago, discussing accountability will come to naught. It is only once international organizations are treated, in law, as the autonomous political actors they are, that discussing their accountability towards third parties has a chance of success.

In addition, at the risk of sounding Weltfremd, much also depends on organizational culture: an organization that internalizes a virtuous mind-set among leadership and staff might be more inclined to behave responsibly than an organization where a ‘Just Do It’ mentality prevails or, worse perhaps, an organizational culture steeped in harshness and rough competition. The point is familiar from studies on business leadershipFootnote ⁶⁴, and may be extended to global governance, including international organizations such as IOM.Footnote ⁶⁵