Bad Cases Make Bad Law, But Good Law Books!

Dr. Marko Milanovic’s book on the Extraterritorial Application of Human Rights Treaties (OUP, 2011), which grew out of his doctoral studies in Cambridge, offers an excellent analysis of the jurisprudence of international and national courts and committees on the extraterritoriality of state obligations in the field of human rights. It is by far the most comprehensive book that has been written on the subject, and I have no doubt that it will quickly become the standard reference text on human rights and extraterritoriality, if it has not already become so. As can be expected, especially by those who have followed Milanovic’s earlier works in the field, he reaches the compelling conclusion that the case law on the extraterritoriality of human rights obligations is hopelessly casuistic and unprincipled, and as a result inconsistent and confusing. Furthermore, he argues that the main ECtHR decision on extraterritoriality – Bankovic v Belgium (2001)– is built on erroneous legal foundations, and runs contrary to previous cases, as well to core human rights values.

 Milanovic is correct in diagnosing most of the reasons for this unhappy state of affairs: The debate over the extraterritorial application of human rights is mired up in a Koskenniemic tension between an ideal (the universality of human rights) and political reality (the principle of effectiveness, which militates against normative overreach). In fact, one can identify a parallel tension at play between the need to ensure effective protection of human rights  (e.g., through eliminating legal ‘black holes’) and the continued commitment to territoriality as an organizing principle of the international legal order, notwithstanding the tenuous connections between borders and human welfare.  A third tension, further complicating the debate on the extraterritoriality of human rights obligations, which Milanovic addresses on a number of occasions, involves the institutional relationship of courts to governments, or law to politics. While the extraterritorial projection of state power is not a new phenomenon in itself, regulating it through legal norms and, even more so, by courts applying international legal norms is a relatively novel development. It is therefore not surprising that courts often treat extraterritoriality as a preliminary jurisdictional question (which Milanovic rightly criticizes as a category error) – jurisdiction to adjudicate being a principal tool that courts employ in order to avoid politically undesirable decisions.

In any event, I agree with Milanovic that judges resent the idea that states should have a free hand to engage in atrocious conduct outside their borders against foreign nationals (and a fortiori against their own nationals). At the same time, the same judges are reluctant to impose on governments (especially their own governments, in the case of national judges, and governments operating in multi-national settings, in the case of international judges) onerous human rights obligations that may complicate what are already difficult international interventions (such as the British occupation of Southern Iraq or the NATO intervention in Serbia in connection with the Kosovo crisis). So, courts try to delineate state obligations in a way that would encompass some, but not too many, extraterritorial exercises of state power, and Milanovic is correct in observing that such a balancing exercise can almost never be based on a coherent application of principles. If a spatial theory of responsibility (effective control over territory) is selected, one is confronted with scenarios involving control over ever-decreasing spaces (a city, a neighborhood, a house, a basement, etc.), which cannot be rationally distinguished from one another. If one adopts a personal theory of responsibility (control over persons by state agents), then again one faces insoluble problems of delineation (e.g., the Al Skeini House of Lords’ unconvincing distinctions between the killing of a person held in UK custody and killings committed by UK forces of five other persons on the streets of Basrah or in their house).

 Milanovic’s book not only exposes the arbitrariness and, outright irrelevance of some of the attempts made by courts to draw bright lines (e.g., Bankovic’s much criticized – and now officially debunked – espace juridique  theory, the legality of the powers exercised by the foreign state or the nationality of the victims); it also offers a new model for delineation of jurisdiction – distinguishing between negative obligations (that apply regardless of jurisdiction, all over the world) and positive obligations (that apply only in territories subject to the state’s effective control). This new model can be couched, more or less, in the language of most human rights treaties, and may be supported by policy considerations: It certainly represents a more principled approach to delineating jurisdiction the than almost all of the existing case law; it also offers a reasonable fit between state capabilities and their human rights responsibilities.

 Still, I fear that, at the end of the day, Milanovic’s own model is also exposed to criticisms of arbitrariness and sub-optimal coverage (and to his credit, he does acknowledge that his model is not foolproof). Most significantly, the distinction Milanovic offers between negative and positive obligations cuts against the increased acceptance of the interdependence (and moral equivalence) of these obligations, and accentuates the already problematic distinction between public and private that permeates international law in general, and human rights law in particular.

 Take, for example, the Herbicide case between Ecuador and Colombia pending between the ICJ. Milanovic suggests that application of his model would lead to the following outcome (assuming that the facts are as argued by Ecuador):

“Colombia would be responsible for violating the rights of Ecuadorian residents adversely affected by its spraying operations, as the operation was conduct [sic] by its own agents or on its own behalf. Colombia would not, however, have the same obligation under human rights treaties with regard to pollution or herbicides used by purely private actors operating within its territory and having effects in Ecuador”. (p. 218)

 But why shouldn’t Colombia protect foreigners from harm caused by private actors situated inside its territory? Certainly, from a victim-protection perspective, individuals should be protected from harm, regardless of whether the harm originates from private and public actors. This basic insight has led, in fact, in the last few decades to the development of doctrines of positive obligations of states under human rights law. Furthermore, in an era where the public/private dividing line becomes more obscure than ever before, putting too much normative reliance on such a distinction could generate serious practical problems and uncertainties. (Would the US, for example, be responsible under Milanovic’s model for not preventing acts undertaken by Backwater Security employees in post-occupation Iraq?). But even from an effectiveness point of view, the distinction offered by Milanovic is not convincing. If Colombia is capable of preventing cross-border harm caused by private Colombian citizens operating inside its territory, then why should it be released from the duty to protect adversely affected foreign individuals? In other words, if Colombia can and must prevent air pollution caused by private actors to individuals inside Colombia, it can also be expected to act when the same private actors, still located inside its territory, harm individuals situated in a neighboring country. Milanovic’s insistence on the location of the victims, and not on the location of the harmful activity as the decisive criteria for application appears, in the circumstances of the case, to be as arbitrary as some of the other distinctions found in the case law he himself criticizes.

 In the same vein, the distinction Milanovic offers between “prophylactic and procedural obligations” attendant to negative obligations (which are mostly second order obligations arising when first order obligations may have breached), such as the duty to investigate deaths caused by state agents, on the one hand, and other positive obligations, such as the duty to protect individuals from other individuals or investigate deaths caused by private actors, on the other hand, cannot be supported from neither a universalist or effectiveness point of view. If states can prevent and investigate both public and private acts of murder (which may have occurred in the same crime scene), why should we require them to comply with positive obligations only with respect to some acts of murder (official murders)?

 At the root of things lie, I think, two parts of the legacy of Bankovic, which Milanovic does not fully reject. First, there remains what I consider to be the most problematic aspect of Bankovic ­– the refusal to allow for partial application of the European Convention. To recall, the ECtHR in Bankovic rejected the notion that the Convention can be “divided and tailored in accordance with the particular circumstances of the extra-territorial act in question”. Milanovic’s position on this aspect of Bankovic is not fully coherent. While his proposal to distinguish between the application of negative and positive obligations implicitly rejects the notion of indivisibility of obligations, he does seem to accept that, for the sake of legal clarity, positive obligations should not be cut up and divided; he therefore insists on effective spatial control as a prerequisite for applying all positive obligations. But this solution is not fully satisfactory either:  If human rights treaties can be divided on the basis of the nature of the obligation, why cannot they be divided on the basis of the content of the obligations, as well? And, more fundamentally, what overriding policy considerations support indivisibility – that is, releasing states from obligations they can carry out, just because there are other obligations, which they cannot fulfill?

In fact, I find it hard to reconcile the notions of indivisibility and extraterritoriality: In a world where governance is still mostly territory-based, the very move to extraterritorial application of governmental power entails a paradigmatic shift in governmental operations, and – by implications – a change in the practical ability of governments to fully implement human rights treaties. Arguably, even in the prototype case for extraterritorial application of human rights law – situations of occupation – not all human rights obligations are relevant or can be realistically applied. For example, it is hard to see how the right to vote for one’s government representatives can be implemented in occupied territories; moreover, some social and economic rights entailing the reorganization of the economy (e.g., introduction of social security or minimum wage) are either impractical or legally impermissible in many, if not all occupation situations. But even if in occupation-like situations there is a strong presumption of full applicability, the farther we move away from occupation situations to smaller units of control, or to other forms of control (personal control or indirect control à la Ilascu, or shared control à la Behrami), the less tenable is the ‘all or nothing’ approach advocated by Bankovic.  And, in fact, much of the case law analyzed by Milanovic – the trial in absentia cases, the passport issuance cases, the extra-territorial search cases etc., cannot be reconciled with the concept of indivisibility. This is because the states whose conduct was reviewed in those cases were hardly in a position to afford the individuals in question all of the protections specified in the human rights treaties to which they are parties. It thus seems to me that considerations of the universal ideal, as well as of effectiveness (in both senses – i.e., that governments shouldn’t be required to provide what they cannot deliver and that human rights should be effectively protected), militate in favor of a flexible or functional approach to jurisdiction: A state should be obliged to protect the human rights of those it is in a position to protect, to the extent that is in a position to do so (subject to limitations I will immediately elaborate on).

Here we come to the second problematic legacy of Bankovic (which Milanovic implicitly accepts, as far as positive obligations are concerned) – the idea that we should be careful not to opt for a “cause and effect” notion of jurisdiction. Indeed, if any state act, which violates a human right, brings the victim harmed by the act within the state’s jurisdiction, jurisdiction becomes a meaningless constraint on the scope of application of human rights law. Although this is a forceful interpretive argument against the over-dilution of jurisdictional conditions, it does not negate, in my view, the introduction of flexible criteria for application that go beyond the existing case law, but do not slide into nothingness – no jurisdictional criteria at all.

First, like Milanovic, I believe that the key aspect of jurisdiction – as the term is used for the purposes of determining the applicability of human rights obligations – is not the actual exercise of rights, but the potential (or functional capacity) to apply them. From this point of view, negative and positive rights do differ, but only for evidentiary purposes:  A violation of a negative obligations shows that jurisdiction had existed (i.e., there has been a capacity to harm), whereas for positive obligations such a proof is found not in their violation but in their actual implementation (i.e., actual protection underscore the capacity to protect). So, in a Bankovic type situation, it is not the act of bombing the TV station in Belgrade which brought the individuals found there under the jurisdiction of the bombing states, but rather the designation of the TV station as a target of attack or, alternatively, the locking in of the fighter planes’ bomb-sights on the target – creating thereby conditions in which a the violation of article 2 of the European Convention could take place. The act of bombing merely illustrated the capacity to exercise power over the victims. At the same time, there was no duty in a Bankovic kind of situation to protect the other rights of the individuals in the TV station, because there was clearly no capacity to do so. So, the individuals in the TV station fell under the jurisdiction of the bombing states for a short period of time only, and with regard to one substantive ECHR right – the right to life (article 2).

Still, I concede that a test of potential or functional capacity to apply some or all of the obligations found in human rights treaties may be too broad a notion of jurisdiction, which may lead indeed, as suggested by Milanovic, to untenable outcomes – primarily, to the imposition of obligations which would be politically unacceptable to states (e.g., an obligation on the US to feed the starving population of North Korea). Milanovic is right in cautioning us against formulating legal standards that deviate too far from our legal (and moral?) intuitions about what states should do (as opposed to what they can do). Still, instead of Milanovic’s model (no limits on negative obligations + positive obligations in areas under effective control), I would opt for a single concept of jurisdiction, applicable both to negative and positive obligations, which centers on the strength of the governmental power that is being applied or can be applied vis-à-vis the individuals in question– a position that I believe find support in the text of HRC General Comment 31 which states: “ State party must respect and ensure the rights laid down in the Covenant to anyone within the power or effective control of that State Party, even if not situated within the territory of the State Party” (para. 10).

So how can one distinguish between exercises of power that entail applicability and those that do not? In a piece I have co-written with Orna Ben-Naftali some eight years ago, we suggested that an extraterritorial projection of power by a state would qualify as encompassing affected individuals within its jurisdiction if the potential impact of the act or policy in question is direct, significant and foreseeable. This standard, whose contents were informed by ‘points of contact’ analysis originating in private international law, is designed to strike a balance between the universality ideal and the reasonable expectations of states as to the legal implications of their action. Note that this proposed test lead to somewhat different outcomes than those deriving from Milanovic’s model: Although our test applies most clearly to situations involving actual or planned governmental acts (negative obligations), it may also cover positive obligations – failures to act which are potentially tantamount to acts in their directness of causation, significant impact and foreseeability (e.g., toleration of private cross-boundary harm in the Colombia/Ecuador case). At the same time, the test we proposed releases states from the indirect harms that may be caused by some of their public actions (e.g., emissions of government factories leading, in the long run, to global warming).

A second restraining notion that I propose for understanding jurisdiction, which is particularly relevant for positive obligations, involves the concept of special power relations (or relative advantage in implementing human rights standards). The general standard that appears to emerge from the case law discussed by Milanovic is that states ought to implement human rights obligations in those situations in which they have a special relationship to the individuals in question that render those states particularly well situated to protect them. Such a relationship may be founded either on strong factual power relations (e.g., effective control over the territory in which the individual resides or physical control over the person), or on a special legal relationship, which generates strong expectations that a particular state would extend its protection over certain individuals. Hence, contrary to Milanovic, I think that the flagship state is particularly well situated to protect the individuals on board its vessels, at least when they sail on the high seas; hence, the state should be legally required to do apply those human rights protections it is able to afford such individuals. In the same vein, the UK should protect the rights of its military personnel in Iraq because it is particularly well situated to do so given its exclusive legal jurisdiction over them.  At the same time, no special factual or legal power relationship exists between the US and the population of North Korea, which would justify holding the former responsible for the human rights of the latter.

Of course, the application of human rights under both of the proposed standards (intensity of potential impact and special power relationship) need not be exclusive – more than one state may impact a certain individual or have a special relationship of power with her (especially, if we allow cutting and dividing human rights treaties).  Furthermore, as Milanovic repeatedly notes throughout his book, having jurisdiction over a certain situation does not entail an absolute obligation to implement the human rights held by the relevant individuals – only a relative obligation to do so. So, Colombia’s duty to prevent cross-boundary harm caused by private citizens or the UK’s duty to investigate inter-Iraqi violence in Basra are only due diligence obligations – no more and no less.

The recent ECtHR Grand Chamber judgment in Al Skeini – issued after the book went to print – provides us with a good opportunity to test the model advocated by Milanovic, and to examine how the alternative tests he and I propose could relate to the most recent statement of the extraterritorial application of the ECHR (found mainly in paragraphs 133-142 of the judgment). Significantly, the Al Skeini judgment internalizes a number of Milanovic’s criticisms of Bankovic (most importantly, it rejects the espace juridique theory), and goes some way in its treatment of the concept of jurisdiction in the direction to which Milanovic points in his book.

First, the Court restates past instances in which control over persons was found to entail the extraterritorial application of the European Convention– exercise of diplomatic or consular authority, exercise of public powers by agreement of the host government and physical power and control over individuals (or ships). Significantly, the Court explicitly reverses Bankovic’s ‘all or nothing’ approach – as far as it is relevant to personal control cases, and states that:

“It is clear that, whenever the State through its agents exercises control and authority over an individual, and thus jurisdiction, the State is under an obligation under Article 1 to secure to that individual the rights and freedoms under Section 1 of the Convention that are relevant to the situation of that individual. In this sense, therefore, the Convention rights can be “divided and tailored” (compare Banković, cited above, § 75)”. (para. 137)

To my mind, all of these cases are indeed instances of intense potential impact or special power relationship, which should have indeed led to the full application of the relevant Convention provisions. (Note that Milanovic ‘s model would apply only negative obligations in these circumstances).

The Court then addresses jurisdiction based on control over territory. Here it accepts the notion of effective control as the basis for full applicability of the Convention (in line with Milanovic’s spatial model). Note, however, that the Court appears to reaffirm there its 2004 Ilascu judgment, implying that a state’s effective control over foreign territory will be determined inter alia in light of “the extent to which its military, economic and political support for the local subordinate administration provides it with influence and control over the region” (para. 138). This flexible approach to what constitutes effective control also sits well, I believe, with the intense potential impact and special power relationship standards for applicability I have identified earlier.

Finally, in applying the existing legal standards to the facts of the case, the Court reaches the unambiguous conclusion:

It can be seen, therefore, that following the removal from power of the Ba’ath regime and until the accession of the Interim Government, the United Kingdom (together with the United States) assumed in Iraq the exercise of some of the public powers normally to be exercised by a sovereign government. In particular, the United Kingdom assumed authority and responsibility for the maintenance of security in South East Iraq. In these exceptional circumstances, the Court considers that the United Kingdom, through its soldiers engaged in security operations in Basrah during the period in question, exercised authority and control over individuals killed in the course of such security operations, so as to establish a jurisdictional link between the deceased and the United Kingdom for the purposes of Article 1 of the Convention (para. 149).

 It is interesting to note that the Court does not explicitly base its decision in this dispositive paragraph on the occupation of Basra by Coalition forces (although this fact is mentioned elsewhere in the judgment) – and the legal obligations (and special legal power relations) that ensue from this state of affairs. It does, instead, emphasize that the UK substituted the authority of the local Iraqi government. In other words, it found that the UK was both capable and particularly well situated to assume the responsibilities of governance vis-à-vis the population of Basra, including the obligations to protect their human rights. On this, Milanovic and I (and the ECtHR) appear to be in full agreement.

 In sum, Milanovic has provided us with an excellent, thought provoking and original book. I fully agree with his analysis of the shortcomings that can be found in the existing case law – bad cases do indeed make bad law. At the same time, I only partly agree with his prescription for a reinterpretation of human rights law – which goes some of the way, but not all of the way towards accommodating principle and politics. This disagreement does not detract, however, from my deep appreciation of the work. So, while bad cases make bad law, they sometime generate (very) good law books.