University of Melbourne Law School Research Series
Last Updated: 23 November 2009
James C. Hathaway[*]
Until the mid-1980s, only refugees could invoke international law to resist
removal to a dangerous country of origin. The evolution
in international law
since that time has been both fast-paced and profound.
This is most clearly true under European human rights law. No less an authority than the House of Lords has declared that the right of non-return extends not only to refugees, but to any person at risk of torture or inhuman or degrading treatment or punishment, and – at least where the risk is clear and extreme – applies also where any of the rights to life, freedom from slavery, liberty and security of person, to protection against ex post facto criminality, the right to privacy and family life, and to freedom of thought, conscience or religion is threatened.
But the dramatic expansion of protection is not limited to Europe. While
less easily enforced than the rules of the European safety
net, the combination
at the global level of an explicit duty of non-return in the United Nations
and of implied duties of non-return grounded in recent authoritative application
of Arts. 6 and 7 of the Civil and Political
establishes a principled limit on the right of the most states to remove a
broadly defined group of at-risk non-citizens from
embryonic support for an expanded duty to protect may be found in the Convention
on the Rights of the
Child and through
invocation of international humanitarian law, specifically common Art. 3 of the
The problem is that none of these new sources of international protection expressly sets out just how members of the broader class of non-returnable persons is to be treated. In contrast to the Refugee Convention, nearly all of which is devoted to defining the precise legal entitlements of members of the protected class (for example, to freedom of internal movement and other civil liberties, as well as to key socioeconomic rights, including to work and to access education), the new protections against refoulement are bare-bones entitlements. Members of the beneficiary class may not be returned to the place of risk, but there is no express duty in international law to provide them with any particular bundle of rights, much less to enfranchise them in the host community.
This situation is grave for the protected persons themselves, and is potentially destabilizing for the governments and communities that host them. The same arguments advanced at the birth of refugee law, namely that the social inclusion of non-returnable persons is both ethically right and socially responsible, have clear contemporary resonance. Perhaps for this reason the European Union’s 2004 Qualification Directive came remarkably close to granting Refugee Convention rights to all persons legally entitled to protection against refoulement and is likely soon to be revised to go even farther in that direction. Canada has already amended its domestic law to give the same rights to refugees and other legally non-returnable persons, with Australia and New Zealand poised to follow suit.
But does international law actually require states bound by one or more duties of non-refoulement beyond that set by the Refugee Convention itself to adopt rights-granting measures of this kind? Is the provision of civil liberties and socioeconomic entitlements to non-returnable persons simply good policy, or is it legally compelled?
Recent scholarship suggests that granting rights to all non-returnable persons is not just advisable, but is already required by international law. First, it is said that even states not bound by relevant conventions are required by customary international law to honor the duty of non-refoulement in relation to a wide-ranging group comprised of both refugees and others facing the prospect of serious harm. Second, it is suggested that all persons who are entitled to protection against refoulement – not just refugees – must, by virtue of a conceptual fusion of the Refugee Convention and other human rights accords, be granted all of the refugee-specific entitlements codified in the Refugee Convention itself.
In essence, under the first claim, the protection of refugees against refoulement ceases to be a matter of treaty-based entitlement. And under the second claim, the specific, treaty-based entitlements of refugees are deemed applicable to all beneficiaries of the duty of non-refoulement, whether refugees or not.
Taken together, the two claims amount to an assertion that there is today a legally binding and universally applicable right to asylum for all seriously at-risk persons. In short, the right to asylum has been leveraged through scholarly analysis despite its express rejection by states.
I believe that the analysis underlying the leveraged right to asylum is conceptually flawed. As I will show, there is no duty of non-refoulement that binds all states as a matter of customary international law and it is not the case that all persons entitled to claim protection against refoulement of some kind are ipso facto entitled to refugee rights. These claims are unsound precisely because the critical bedrock of a real international legal obligation – namely, the consent of states evinced by either formal commitments or legally relevant actions – does not yet exist.
The leveraged right to asylum is attractive because it allows scholars simply to ordain law, rather than having to work to create a renewed protection architecture that convinces states that extant duties can be reconciled to national self-interest. But the effort to rebuild, rather than simply to ordain, is required precisely because rhetorical claims standing alone will not serve as a meaningful constraint on the behavior of states.
The First Claim: The Duty of Non-Refoulement Binds All States
The essence of the first claim is that because an express or implied duty of
non-refoulement is recognized in the various treaties I have mentioned,
it is now the case that all states – whether bound by a relevant
treaty or not – are legally obligated to honor the duty of
non-refoulement vis à vis not only any refugee, but also in regard
to any potential victim of torture, cruel or inhuman or degrading treatment
punishment (“torture”), as well as most persons facing risk to
“life, physical integrity, or
The claim is that at least this one critical refugee right inheres in all
persons who are in fact refugees or who face another serious
human rights risk,
regardless of treaty accession. How was this conclusion reached?
In a study commissioned and championed by the United Nations High Commissioner for Refugees (UNHCR), Eli Lauterpacht and Daniel Bethlehem invoke the decision of the International Court of Justice (ICJ) in the North Sea Continental Shelf Cases for the view that treaties “may influence the creation of... a rule of custom.” They argue that because the treaty-based norm of non-refoulement is of norm-creating character, enjoys widespread and representative state support, and has stimulated consistent relevant practice, “non-refoulement must be regarded as a principle of customary international law.” If sound, this analysis means that the duty of non-refoulement is no longer merely a matter of treaty-based obligation (applicable only to state parties) but instead now binds all states, including those which have never signed on to a relevant treaty.
The basic notion that customary law may emerge from a treaty-based norm is well-accepted. At least since the Asylum Case, it has been recognized that the tree of customary international law can grow from the acorn of specific treaties. But the role of the treaty-based norm is essentially auxiliary: it crystallizes the content of the putative norm and provides a context within which the two essential elements of a customary norm – opinio juris and consistent state practice – can be located. In the case of the putative customary duty of non-refoulement, these two essential requirements for the emergence of customary international law are not made out.
To begin, is there opinio juris sufficient to justify the putative norm? The rigid traditional understanding of opinio juris sive necessitatis – requiring that the observed uniformity of practice be a consequence of a sense of legal obligation – has, of course, given way to the less demanding requirement “of an express, or most often presumed, acceptance of the practice as law by all interested states.” It is sufficient to show that states presently regard the putative norm as legally compelled, even if their concordant actions in keeping with the norm were not induced by a sense of legal duty. There is moreover good authority that opinio juris can be shown in many different ways. In its Nicaragua decision, for example, the ICJ held that “opinio juris may, though with all due caution, be deduced from, inter alia, the attitude of... States towards certain General Assembly resolutions... support of [regional conference] resolutions... [and] statements by State representatives.” The views of most scholars are similarly inclusive.
Despite this very flexible approach to the material basis for identification of opinio juris, the specific facts relied upon by Lauterpacht and Bethlehem fall short. They ground their claim of opinio juris for a universally binding duty of non-refoulement on a combination of three indicia: first, the “near-universal acceptance” of a non-refoulement duty in various UN and regional treaties; second, the unanimous adoption by the General Assembly of the 1967 Declaration on Territorial Asylum; and third, the absence of express opposition to the principle of non-refoulement by the states which neither signed a relevant treaty nor were present in the General Assembly when the 1967 declaration was adopted.
The primary portion of the claim is substantively rickety. For a single rule of customary international law to emerge, the indicia of opinio juris must clearly relate to the same putative rule. In contrast, Lauterpacht and Bethlehem weave together disparate bits of opinio juris arising from distinct treaties dealing with distinct issues to locate opinio juris for a principle that is more comprehensive than any of the underlying commitments. Specifically, they argue that because all but nineteen UN member states “... participat[e] in some or other conventional arrangement embodying non-refoulement...” – that is, they have all agreed to be bound by at least one of Art. 33 of the Refugee Convention, Art. 3 of the Torture Convention, Arts. 6 and 7 of the Civil and Political Covenant, or by a comparable provision under a relevant regional treaty – it is now possible to conclude that there is a sufficiently widespread and representative opinio juris for an overarching principle that “... non-refoulement must be regarded as a principle of customary international law.”
The incongruity of the claim arises from the fact that non-refoulement is merely a means to a protection end. The means itself can only be the subject of general acceptance within a particular context. That is, the assertion that all states accept the duty of protection against refoulement assumes some agreement about the circumstances in which the duty is owed. Yet there is no such agreement, since the evidence of opinio juris relied upon by Lauterpacht and Bethlehem sometimes relates to persons who have a well-founded fear of being persecuted; in other cases, to persons at risk of torture; and in still other circumstances, to persons at risk of other forms of human rights abuse. There is, in short, no common acceptance of the duty of non-refoulement related to any particular class of persons or type of risk, much less to their combined beneficiary class.
By way of analogy, one might consider the claim that there is opinio juris to support an international legal duty to issue an injunction. At one level, it is true that the courts of virtually all states do, in fact, authorize injunctive relief in at least some circumstances. Yet it would be meaningless to claim a normative consensus on a duty “to issue injunctions” since there is no substantive accord on the circumstances in which the remedy is to be granted. The argument for opinio juris in support of a general duty of non-refoulement is similarly flawed. Lauterpacht and Bethlehem’s assertion of agreement sufficient to count as opinio juris is a thinly veiled, cobbling together of disparate commitments with only the veneer of a remedial mechanism – non-refoulement – in common. With no substantive commonality to the obligations agreed, no general opinio juris can be derived.
The second form of evidence of opinio juris relied upon by Lauterpacht and Bethlehem, the unanimous adoption by the General Assembly of the 1967 Declaration on Territorial Asylum, does have a common substantive core. Unfortunately for their project, the common core is limited to persons seeking “asylum from persecution,” a group far smaller than said by them to benefit from the customary norm. Equally fundamental, General Assembly resolutions cannot be relied upon in abstracto as evidence of universal opinio juris. As the ICJ observed in Nicaragua, the opinio juris is instead to be deduced from “the attitude of... States towards certain General Assembly resolutions (emphasis added).” The Court noted that while General Assembly resolutions may be the basis for opinio juris, they have to be considered “in their totality.” A critical part of that totality is the fact that a United Nations conference convened in 1977 with the specific intention of transforming the 1967 declaration into binding law was a failure. Lapenna notes that “the Committee met for [more than] four weeks, and only three of the ten articles of the experts’ draft were discussed and voted on... [T]he preoccupation of the majority of the states was that of safeguarding, to exasperation point, the sovereign right of a state to grant asylum.” There has moreover been no subsequent effort to revisit the asylum convention project. To rely on the 1967 asylum declaration as an indication of state parties’ acceptance of a comprehensive duty of non-refoulement – much less to isolate the nineteen abstaining countries and deem their failure to protest to be implied support – is disingenuous given the totality of the evidence of state attitudes.
The more plausible basis for General Assembly-based opinio juris – ironically, not invoked by Lauterpacht and Bethlehem – is the line of subsequent General Assembly calls to respect the duty of non-refoulement, often said to apply to “asylum seekers” as well as to the arguably more constrained category of refugees. While not as specific as the beneficiary category contended for, the regularity of the endorsement of non-refoulement in the General Assembly is noteworthy, and goes some distance in support of the claim that there is opinio juris for a duty of non-refoulement owed to more than just Convention refugees.
The challenge, though, is that General Assembly resolutions are merely one factor to consider in the assessment of opinio juris. They must be weighed-up against contrary indications, in particular those emanating from states not already bound by a treaty-based duty of non-refoulement. Apart from the failure of (and failure to resuscitate over the ensuing forty years) the territorial asylum initiative described above, the major contraindication is the persistent refusal of most states of Asia and the Middle East to be formally bound by the asserted comprehensive duty of non-refoulement. While such states have often agreed to admit refugees and other human rights victims, there is no evidence that whatever openness they have shown – often partial, and usually highly conditional – has been influenced by a sense of legal obligation (rather than, for example, by political or economic calculus, social or cultural affiliation, or a sense of moral responsibility). A former Chief Justice of India, for example, affirmed that while courts in his country “have stepped in” on occasion to prevent refugee deportations, “... most often these are ad hoc orders. And an ad hoc order certainly does not advance the law. It does not form part of the law, and it certainly does not make the area clear.” As the ICJ noted in the North Sea Continental Shelf Case, such actions do not support a finding of opinio juris.
The persistent reluctance of the majority of states in Asia and the Middle East to embrace a comprehensive legal duty to protect refugees and others against refoulement is problematic for a second reason. Customary international law formation sensibly gives particular attention to the views of states “specially affected” by the phenomenon sought to be regulated. With Asia and the Middle East hosting the majority of refugees in the world yet failing clearly to affirm a duty to protect, the assertion of universal opinio juris based on General Assembly resolutions is thus especially fragile.
To be clear, I recognize that when a treaty-based norm stimulates a broadly embraced sense of obligation (in particular, among non-party states), opinio juris in support of a cognate customary international legal obligation may emerge. But there is no basis to assert that just because most countries have accepted something that may broadly be termed a non-refoulement obligation, applying to at least some kinds of cases, and in at least some contexts, it can now be concluded that there is a universally applicable duty of non-refoulement owed to the combined class of all refugees and other persons at risk of significant human rights abuse. Much less can opinio juris be located in General Assembly resolutions considered in isolation from the more general context of state attitudes towards the putative norm, in particular the attitudes of states specially affected by refugee flows.
But even if opinio juris could be located, the next question that must be addressed is whether there is evidence of consistent state practice that aligns with the putative norm (the second essential element for establishment of a customary law). Sadly, there is in fact very significant empirical evidence that undermines the claim of state practice in conformity with a broad-ranging and universally applicable duty of non-refoulement. For example, a recent survey of the 52 countries hosting the largest number of refugees found that 35 such states – that is, more than 2/3 of the states examined – had committed “some” (11), “significant” (10), or “severe” (14) acts of refoulement or comparable physical endangerment. In nearly a quarter of the countries evaluated, the risk was adjudged to be intensifying over time. This data moreover exists against a long-standing and extensive pattern of refoulement across the world, including blatant refusals to allow refugees to access state territory; turn-back policies implemented by the closure of borders to refugees; the construction of blunt physical barriers to prevent the entry of refugees; summary ejection of refugees able physically to cross a border; removals ordered without access to a procedure to verify refugee status; expulsions consequent to the failure to ensure even basic procedural safeguards in the assessment of refugee status; disguised removals under the rubric of “voluntary” repatriation; arms-length maneuvers to repel refugees in areas of arrogated jurisdiction beyond a state’s borders; and the establishment of non-entrée legal regimes that prevent refugees from even reaching the point of being able to present their case for protection.
There is, in short, pervasive – perhaps even dominant – state practice that denies in one way or another the right to be protected against refoulement. These surveys of state practice are moreover restricted to the comparatively well-protected category of “refugees”; it is likely that the refoulement of the broader categories of human rights victims claimed by Lauterpacht and Bethlehem to be part of the beneficiary class of the customary norm is even more pervasive. How, then, can it be argued that there is relatively consistent state practice in conformity with the putative universal duty to protect refugees and other human rights victims against refoulement?
First, some argue that the depth and consistency of state practice required for the establishment of customary international law should not be overstated. So long as respect for non-refoulement remains the norm, it is suggested that the state practice requirement is met. Second, and impliedly allowing for the inadequacy of an empirical record of concordant practice, there is authority for the view that so long as there is an effort to justify acts of refoulement as permissible exceptions to the alleged norm, practice that is on its face violative of the norm is in fact supportive of it. And third and most significantly, it is claimed that while state practice is required, real state action on the ground may be overcome by alternative “practice” in the form of verbal commitments to protect refugees against refoulement. I will consider each of these claims in turn.
First, what of the view that the depth and consistency of state practice required for the establishment of customary international law should not be overstated? There has certainly been a trend in the ICJ jurisprudence to soften the standard of uniformity required. The 1950 Asylum decision spoke of “constant and uniform usage,” the 1969 North Sea Continental Shelf case stated the test as “extensive and virtually uniform” practice, while the Nicaragua decision of 1986 noted that “absolutely rigorous conformity” is not required. It is thus easy to see why scholars are disinclined to set an overly demanding threshold of consistency of state practice. Brownlie, for example, opines that consistency of state practice “is very much a matter of appreciation.”
That having been said, there is little doubt that clearly predominant global practice remains a requirement for the establishment of a customary legal duty. The ICJ’s exhortation in the Asylum decision that “fluctuation and discrepancy” in practice undermines the argument for custom is a helpful, and surely not overstated, indicator of the circumstances in which consensus through action is simply not present. While those seeking to downplay the relevance of practice often rely on the Court’s statement in Nicaragua that custom can arise despite “not infrequent” inconsistent practice, this obiter dictum must be balanced against the same judgment’s insistence that a “settled practice” be identified. More specifically, as Villiger writes, state practice for a customary norm binding all states must at least be “general” in the sense “that common and widespread practice among many States is required. While universal practice is not necessary, practice should be ‘representative,’ at least of all major political and socio-economic systems.”
Assessed against even this relatively low benchmark, the case for identification of consistent state practice in line with a broadly inclusive duty of non-refoulement fails. Not only is there a record of refoulement in the majority of the states hosting most of the world’s refugees – and becoming more common in roughly a quarter of them – but there is a clear geopolitical skew to the pattern of non-compliance, with half of the major hosting countries with the worst records on refoulement located in Asia or the Middle East. To suggest that there is anything approaching a “settled practice” of non-refoulement, much less a settled practice that is geopolitically inclusive, defies analysis.
Nor is the case for a settled practice in line with the duty of non-refoulement assisted by a second argument, namely that breaches can sometimes support a finding of consistent state practice. The ICJ’s Nicaragua judgment, generally regarded as the most authoritative statement of this rule, is at pains carefully to explain the basis for its holding that “instances of State conduct inconsistent with a given rule... treated as breaches of that rule” contribute to a finding of consistent state practice in support of the norm:
If a State acts in a way prima facie incompatible with a recognized rule, but defends its conduct by appealing to exceptions or justifications contained within the rule itself, then whether or not the State’s conduct is in fact justifiable on that basis, the significance of that attitude is to confirm rather than to weaken the rule.
In that case, the question was whether instances of foreign intervention in support of an internal opposition group espousing “worthy... political or moral values” – at least prima facie in breach of the putative norm of non-intervention – had been defended on the basis of justifications or exceptions said to be part of the putative norm itself. The manner in which the argument was rejected is instructive:
... [T]he Court finds that States have not justified their conduct by reference to a new right of intervention or a new exception to the principle of its prohibition. The United States authorities have on some occasions clearly stated their grounds for intervening in the affairs of a foreign State for reasons concerned with, for example, the domestic policies of that country, its ideology, the level of its armaments, or the direction of its foreign policy. But these were statements of international policy, and not an assertion of rules of existing international law. In particular, as regards the conduct towards Nicaragua which is the subject of the present case, the United States has not claimed that its intervention, which it justified in this way on the political level, was also justified on the legal level, alleging the exercise of a new right of intervention...
Much the same concerns arise from an examination of state practice of
refoulement. To begin, most instances of refoulement appear not
to be justified at all – they simply
occur. And where
an effort to justify refoulement is made, states tend to offer only blunt
and unsubstantiated assertions that those seeking protection are not refugees or
political cost of protection is too
high. There is,
in short, rarely an effort made to justify turn-backs and other acts of
refoulement by reference to the norm of non-refoulement itself,
much less by arguing the applicability of the internal limitations to that
duty. As such,
inconsistent practice is just that – inconsistent, and hence at odds with
the assertion of a customary legal duty.
This analysis leaves us, then, with one final argument in support of state practice sufficient to ground a broad duty of non-refoulement in customary international law. The essence of the argument is that a very broad reading of “state practice” is justified under which words alone may amount to “practice.” The proponents of this position look to many of the same statements relied upon to show opinio juris as the relevant practice in support of the norm, and thereby arrive at the conclusion that consistent state “practice” can be located despite the evidence of non-conforming “practice on the ground” previously identified.
It is in regard to this issue that the rules of customary law formation are most contested. As Kammerhofer explains, there is a tendency among many academics to define “practice” in a way that obviates the distinction between practice and opinio juris:
Behind the apparent dichotomy of “acts” and “statements” lies a more important distinction: that between one argument that sees practice as the exercise of the right claimed and the other that includes the claims themselves and thus blurs the border between the concept of “state practice” and “opinio juris.”
This is indeed the nub of the controversy: despite the continued insistence of the ICJ that there are two, not one, essential elements to the formation of customary international law, there seems to be a determined academic effort to downplay that requirement. The Final Report of the International Law Association (ILA) Committee on Formation of Customary (General) International Law provides a classic example of this propensity to confuse:
The Court has not in fact said in so many words that just because there are (allegedly) distinct elements in customary law the same conduct cannot manifest both. It is in fact often difficult or even impossible to disentangle the two elements.
The language used is quite extraordinary: note that the ILA does not
say that the International Court of Justice has held that both elements
of custom may be manifested by the same, presumably purely verbal, evidence, but
rather simply that it “has
not... said in so many words” that it
This cautious, if convoluted, framing is warranted given the actual state of ICJ jurisprudence. The decision in Nicaragua, while often cited as the leading source of the notion that words alone can constitute state practice, did not actually reach that conclusion. The focus of the dispute was whether there was a customary norm prohibiting the threat or use of force against the territorial integrity or political independence of a state that parallels the treaty-based rule in Art. 2(4) of the UN Charter. The Court was insistent that a customary norm could arise only upon proof of “the actual practice and opinio juris of States.” For good measure, it added:
The mere fact that States declare their recognition of certain rules is not sufficient for the Court to consider these as being part of customary international law... [I]n the field of customary international law, the shared view of the Parties as to the content of what they regard as the rule is not enough. The Court must satisfy itself that the existence of the rule in the opinio juris of States is confirmed by practice.
The common confusion about just what the Court decided arises from the fact
that it took what can only be described as a fairly slipshod
approach to the
assessment of state practice before focussing on the issue of opinio
Implicit in its analysis that “[i]t is not to be expected that in
the practice of States the application of the rules in question
should have been
is too high a standard, and that prima facie violations justified by
reference to the rule itself “confirm rather than weaken the
an assumption – admittedly, an empirically suspect
– that one could reasonably assume evidence of relatively
consistent state practice of non-intervention other than as authorized
Because the parties chose not to contest the issue of state practice, the Court
understandably focussed its analysis on the opinio juris question,
finding (as previously
noted) that a
wide-ranging set of verbal acts could give rise to opinio
The Court is, however, explicit that these verbal acts are approved strictly as forms of opinio juris, not state practice. As such, and despite the failure of the Court clearly to interrogate the state practice dimension of the claim, it is disingenuous to suggest that its lack of precision in this regard amounts to an endorsement of a new theory of customary international law formation in which state practice is rendered virtually identical to opinio juris. If this had been the Court’s intention, why would it have been at such pains to confirm the traditional two-part test and address the sufficiency of imperfect state practice?
It follows that the notion that verbiage without concordant state practice gives rise to customary law is at best de lege ferenda rather than settled law. Four main arguments are posited in favor of this approach: plain meaning allows it; it avoids a detrimental reliance concern; states want it; and it promotes international order and human values.
On the first point, Villiger argues that “the term ‘practice’ is general enough – thereby corresponding with the flexibility of customary law itself – to cover any act or behaviour of a State, and it is not... entirely clear in what respect verbal acts originating from a State would be lacking...” While linguistically plausible and with at least some support in the jurisprudence, the double-counting of the same words as both opinio juris and relevant practice is difficult to square with the ICJ’s continued insistence on both evidence of state practice and opinio juris. If words evincing acceptance as law are the essence of opinio juris, a court inclined to view words as sufficient state practice ought simply to have dispensed with the dual requirement – which the ICJ has not.
A second argument for treating words alone as practice advanced by Villiger is grounded in the importance of avoiding detrimental reliance. He writes that “whatever a State feels or believes when making a statement, at least other States may come to rely on this statement, and the original State may even be estopped from altering its position.” This is a circular argument, since if it is clear that only practical actions “on the ground” count as relevant state practice, then there is no reasonable basis for other states to put stock in pure statements, thus disposing of the risk of detrimental reliance.
Underlying this argument, though, is a third point. Oscar Schachter has argued that in at least some circumstances states seem to want statements standing alone to be treated as practice relevant to the formation of custom. In Schachter’s opinion, “... in the contemporary international milieu governments have felt a need for a new law which, for one reason or another, could not be fully realized through multilateral treaties... For one thing, the processes of treaty negotiation are often slow and cumbersome... In these circumstances, it has been natural for States to turn to law-declaring resolutions of the General Assembly.” There is in Schachter’s view implied consent for treating at least this one form of “words alone” – namely, law-declaring resolutions adopted unanimously or without significant dissent – as what amounts to instant customary international law.
This is, of course, a much narrower point than the general argument in favour of treating words generally as state practice. And Schachter is far from alone in wishing to see at least some resolutions of international organizations, in particular of the General Assembly, treated as a special example of “state practice.” As Jennings and Watt opine, “the concentrations of state practice now developed and displayed in international organizations and the collective decisions and activities of the organizations themselves may be valuable evidence of general practice accepted as law in the fields in which those organizations operate.” But there are nonetheless several concerns.
First, the fact that the General Assembly is explicitly denied the right to engage in general lawmaking activities should give pause before attributing special lawmaking force to its resolutions. Second, it seems odd to argue that governments have effectively consented to use of the General Assembly as a lawmaking forum in order to overcome the (presumably overly demanding) procedural requirements of lawmaking by treaty when those same governments have declined either to amend the rules of treaty-making or the Charter to provide for the speedy process Schachter assumes they want. And finally, where precisely is the evidence that states – rather than scholars – really want a speedy, less formal lawmaking process? The only example Schachter provides in support of his thesis is the adoption in 1946 of resolutions condemning genocide as a crime and approving the Nuremberg Principles. Both the paucity of examples and the fact that the instances cited by Schachter led to subsequent codification in treaty form suggest that support for the “states want it” thesis is modest at best.
This leaves us with one final argument for treating verbal statements as practice, namely that the world needs a lawmaking process capable of generating results in some core areas, even if state consent cannot be located through one of the general modes (including via consistent practice in the case of custom). In advancing this thesis, Schachter forthrightly acknowledges its instrumentalist tenor, writing that “[t]he problem of inconsistent practice (ie. violations) comes up sharply in respect of declared norms of human rights... In the face of these facts, it is hard to conclude that the declared norms are confirmed by general and consistent practice.” And he is equally candid in noting that “[m]ost international lawyers seek to minimize the violations by emphasizing strong verbal condemnations and denials... [But] [t]he notion that contrary practice should yield to opinio juris challenges the basic premise of customary law.”
Schachter’s solution is to endorse that contradiction in relation to only a subset of customary lawmaking – namely, where putative norms “are strongly supported and important to international order and human values.” He argues that in this context “the norm has to be maintained despite violations” because “they are brittle in the sense that violations are likely.” A more systematized version of this approach is offered by Frederic Kirgis, who asserts that the two elements of customary lawmaking – opinio juris and consistent state practice – should be viewed “not as fixed and mutually exclusive, but as interchangeable along a sliding scale”:
The more destabilizing or morally distasteful the activity – for example, the offensive use of force or the deprivation of fundamental human rights – the more readily international decision makers will substitute one element for the other, provided that the asserted restrictive rule seems reasonable.
Despite the fact that Kirgis speaks of what international decision makers do,
his analysis actually relies on only the Nicaragua case to support the
claim that “a clearly established opinio juris establishes a
customary rule without much (or any) affirmative showing that governments are
consistently behaving in accordance with
For reasons previously given, this is not in my view an accurate interpretation
of the Nicaragua
case. But even
if advanced simply as a thesis de lege
there are good reasons not to endorse the proposed instrumentalist “gloss
the duty to show relatively consistent state practice in support of the putative
Most fundamentally, this view of custom is a disingenuous circumvention of the requirements of lawmaking by treaty. If words alone are to evince state consent to be bound, then those words are required to be formalized as treaty. To treat a wide variety of words uttered in less exacting circumstances not simply as opinio juris but as binding in and of themselves would, as Kelly rightly asserts, be to “constitute a new legislative form of lawmaking, not customary international law based on state behavior accepted as law.” Proponents of an exaggerated definition of state "practice" deny the most elementary distinction between treaties and custom: custom is not simply a matter of words, wherever or by whomever uttered, but is a function of what is happening in the real world. Custom, as distinguished from treaty, is about negotiation via practice. The effective obliteration of the consistent practice requirement advocated by many scholars is thus conceptually flawed. As Wolfke has acerbically observed, “repeated verbal acts are also acts of conduct... but only to customs of making such declarations... and not to customs of the conduct described in the content of the verbal acts.”
This is not a purely formalist point. As the huge variation in theories of which words count as practice makes clear, the risk of subjectivity and political distortion inherent in the transmutation of words into practice is extreme. Kelly rightly points to the likelihood of cultural bias in the selection of which norms are “important to international order and human values,” “important [to] the common interests of states or humanity,” or which address concerns that are “destabilizing or morally distasteful”:
Powerful states use “non-empirical” customary international law to justify the exercise of power without actual acceptance. Environmental and human rights activists, on the other hand, envision customary international law as an instrument for progressive change... [Customary international law] is an inapt instrument for all of these uses. The clever use of arbitral decisions, general dicta from a few ICJ cases, the glorification of general and ambiguous non-binding instruments, or the reconceptualization of customary international law do not establish either requirement of customary law. Custom takes its authority from the belief in the normative quality of resolved experience, not the manipulation of legal instruments.
In sum, “[t]his impressionistic disarray allows the scholar, advocate,
or judge in the few cases that are adjudicated to subjectively
arrive at a
conclusion affected by normative predilection. The customary international law
of human rights is a product of the normative
perspectives of academics and
advocates practicing human rights law, not the social facts of states accepting
Given the inherent subjectivity of treating some, but not all, words as customary law without need for concordant practice, it should come as little surprise that relevant assertions of customary duty rarely attract compliance by states. It is surely right that “[t]he less powerful nations... would be unlikely to accept the ‘claims’ approach of D’Amato or the New Haven school because it would diminish their role in law formation,” even as the view favored by many in the less developed world that “the accumulation of non-binding international instruments creates binding legal obligations is not one which is widely shared by [developed] states and has been specifically rejected by the United States.” This, it seems to me, is the critical answer to scholars such as Schachter who argue for the revaluation of words as practice based on the need to secure critical social ends. If compliance is not in practice advanced by the assertion of words alone as customary international law – and there is little evidence that it is – then on what basis does the appeal to necessity really stand? And if the alleged necessity really does exist in the context of a shared assumption of critical need – as most theorists assume it should – then there will in any event be little difficulty proceeding to a treaty to concretize that agreement.
The Second Claim: Non-Refugees are Entitled to Refugee Rights
Jane McAdam’s pioneering study, Complementary Protection in
links neatly to Lauterpacht and Bethlehem’s analysis. McAdam’s
thesis is that all persons who are entitled to be protected
refoulement are – despite their non-refugee-status – entitled
to the same rights as refugees admitted pursuant to the requirements
To be clear, McAdam’s argument is not simply that
non-returnable persons are entitled to all generic, internationally recognized
human rights. Her claim is specifically
that all persons who are non-returnable
under international law benefit from the specific entitlements which the Refugee
grants to refugees who satisfy the refugee definition set in Art. 1
McAdam’s claim at times appears to be (appropriately) aspirational. She is clearly correct that there is a “protection gap” arising from the fact that most of the new duties of non-refoulement have simply been read into treaty law by authoritative interpretation of the supervisory bodies. In contrast to the Refugee Convention’s explicit design as an instrument to codify the rights of its beneficiary class, the incremental and opportunistic way in which broader duties of non-refoulement have generally evolved provided no comparable opportunity to secure clear agreement on the rights of the expanded class of persons entitled to protection against refoulement. Non-removable non-refugees have thus been forced to rely on generic (and hence often insufficiently attentive) rights set out in general human rights law. McAdam acknowledges that customary law has not intervened to fill this void, forthrightly conceding that state practice does not presently support the attribution of refugee-specific rights to other persons benefitting from protection against refoulement. She points instead to the comparability of needs between non-removable, at-risk, non-citizen refugees and non-removable, at-risk, non-citizen “others.” McAdam encourages us to recognize the flexibility of the Refugee Convention’s rights regime, noting that “... there is nothing intrinsic in the Convention regime that prevents its extension to persons outside the article 1(A)(2) definition...” There is, in my view, much force to this argument, at least seen as a sensible policy option.
Indeed, I believe there is much to commend a second and more legally aggressive argument briefly alluded to by McAdam. Moving beyond the purely normative, she points to the utility of non-discrimination law as the basis for compelling states to grant non-removable “others” the same rights granted to non-removable refugees. Arguing that “... there is no legal justification for distinguishing between the status of Convention refugees and beneficiaries of complementary protection,” McAdam asserts that “[t]o invoke the Convention refugee definition as intrinsically and exclusively legitimate in giving rise to a privileged alien status is... both historically inaccurate and legally flawed.” Noting that non-discrimination law “permits distinctions between aliens who are in materially different circumstances, but prohibits unequal treatment of those similarly placed,” she neatly sets the stage for invocation of the broad-ranging duty of equal protection of the law, especially that set by Art. 26 of the Civil and Political Covenant. Unless the differential allocation of rights between refugees and other non-removable, at-risk, non-citizens is demonstrably “objective and reasonable,” the same rights must be extended to both groups. As such, both in state parties to the Refugee Convention and in countries that grant refugees preferred rights in practice, the duty of equal protection is in my view a powerful basis upon which to assert the need to enfranchise the broad category of persons in receipt of protection against refoulement with refugees.
But McAdam’s claim is neither simply normative nor based on equal protection obligations. Regrettably in my view, she insists that there is a present legal obligation to assimilate refugees and other beneficiaries of protection against refoulement for purposes of rights entitlement because the recognition of non-refugee-specific duties of non-refoulement amounts to an indirect amendment of the scope of the Refugee Convention. She writes:
... [I]nstead of the Convention’s terms being formally expanded by a Protocol or an amendment to the text itself, ... the development of human rights-based non-refoulement has extended eligibility for protection, while the Convention may be appropriately viewed as articulating the resulting status.
As a specialist human rights treaty comprising one part of a holistic human rights regime, it is argued that the Convention’s application has been extended through the expansion of non-refoulement under human rights law (and, by analogy, to protection granted in accordance with humanitarian and international criminal law), rather than by the conventional means of a Protocol... Since the scope of non-refoulement has been broadened by subsequent human rights instruments, this necessarily widens the [Refugee] Convention’s application.
I believe this analysis to be in error.
Going even farther than the Lauterpacht and Bethlehem analysis, McAdam accords a reified place to the duty of non-refoulement. Her premise is that the Refugee Convention is essentially a treaty concerned to identify persons who should be granted protection against refoulement and then to define the rights that attach to persons in receipt of protection against refoulement. Under this rubric, since refugees are only a part of the “non-refoulement-acquiring class” which is in her view the basis for accessing Refugee Convention rights, refugees can receive no more rights than any other beneficiary of protection against refoulement.
But the Refugee Convention is not an instrument that is organized around granting rights to a beneficiary class defined by the duty of non-refoulement. Codification of the duty of non-refoulement was actually far from the core of the Refugee Convention’s purposes; indeed, as initially proposed, Art. 33 would have applied only to refugees arriving with pre-authorization in a state party. As a matter of historical fact, there is no basis to suggest that the Refugee Convention exists to delineate the entitlements of persons granted protection against refoulement.
And even if the Refugee Convention were a treaty intended to define the rights of the beneficiaries of non-refoulement (rather than about the rights of refugees), how does one amend the express beneficiary class of the treaty by stealth? Would it really follow that the express scope of a treaty concerned to provide rights to the beneficiaries of protection against refoulement automatically expands to embrace persons granted comparable protection under other instruments, or under customary international law? Given the clarity of rules about the amendment of treaties, how exactly can it be that “the [broadened] scope of non-refoulement [under] subsequent human rights instruments,... necessarily widens the [Refugee] Convention’s application,” as McAdam suggests?
Her theory of indirect amendment is that “[t]he Refugee Convention provides the clearest statement of international law’s treatment of persons in need of international protection and, as such, this treaty may be seen as providing the status for a more broadly constituted notion of ‘refugee’ (emphasis added).” The Refugee Convention amounts, in McAdam’s view, to “... a form of lex specialis (specialist law) for all those in need of international protection, and provides an appropriate legal status irrespective of the source of the State’s protection obligation (emphasis added).” What is really being said here?
Assuming that McAdam is right (as I believe to be the case) that the Refugee Convention’s rights regime is “the clearest statement” of duties owed to aliens in need of protection, why then does it follow – “as such,” to use McAdam’s language – that it defines the rights of persons to whom it does not textually apply? If the question were whether this would be desirable, the answer is likely “yes.” But McAdam’s is not a mere normative claim; rather, she states it as a legally required conclusion based on no more than the principled logic of the refugee rights list being a good fit with the needs of others. Yet as she forthrightly concedes, not even the regime for stateless persons, drafted contemporaneously with the Refugee Convention, grants all of the same rights to that group as are bestowed upon refugees. To insist nonetheless that the beneficiary class for refugee rights has been de jure expanded to include all those protected against refoulement, whether refugees or not – with no argument based on treaty amendment or on the rise of either a customary or general principles norm – is legally impossible.
Nor is the argument assisted by McAdam’s appeal to the notion of lex specialis. This general principle of international law exists primarily to resolve a conflict between competing international norms; it is also sometimes invoked also to justify reliance on specialized norms to interpret the scope of more general rules. But neither the primary nor secondary meaning of lex specialis provides a legal basis for extending a treaty’s beneficiary class to embrace persons outside its textual ambit.
Rather, the core meaning of lex specialis is that where two rules of international law – one specific, one more general – deal with the same subject matter, the more specific rule governs in the event of conflict. Lex specialis does not require that general rules be ignored where they can be applied without infringing the specific rule; to the contrary, “[f]or lex specialis to apply... there must be some actual inconsistency between [the two rules], or else a discernible intention that one provision is to exclude the other.” Conde provides a helpful example:
... [F]reedom of religious expression (“manifestation of religion”) can be considered lex specialis of the norm of freedom of expression. It carves out a particular area of a more general subject for special treatment. It is usually used in the interpretation of treaty norms as a rule that states that a specific rule will always overrule a general rule covering the same subject.
If therefore a treaty provision on religious freedom were framed in absolute
terms whereas another treaty on freedom of expression
in general were framed
with permissible limitations, a state party to both treaties would be obliged to
respect religious freedom
without reliance on the limitations allowed under the
general accord. But this primary understanding of lex specialis clearly
does not support the view that non-refugees are entitled to refugee rights by
virtue of the similarity of their predicament.
Because there is simply a legal
void to be filled in relation to non-refugees, there is no conflict of rules
that lex specialis can assist to resolve.
The secondary role of lex specialis is similarly irrelevant to
McAdam’s argument. In addition to defining the “trump” in the case of legal conflict, lex specialis may be invoked as an interpretive aid, most commonly to assist in the construction of a general provision in relation to a matter also governed by a more specific norm. In the Nuclear Weapons decision, for example, the International Court of Justice invoked lex specialis to require that the meaning of the (general) provision in the Civil and Political Covenant’s right not to be arbitrarily deprived of one’s life should be construed – in the context of armed conflict – in a way that takes account of the (more specific) provisions of the simultaneously applicable rules of international humanitarian law. And in the Court’s opinion in the Israeli Wall case, lex specialis was again relied on to compel assessment of the legality of the wall not only by reference to the (general) provisions of human rights law, but also taking account of the (more specific) rules of international humanitarian law.
While not resolving a conflict in the same direct way as it does when playing its primary “trump” role, the interpretive variant of lex specialis promotes the same general end – the avoidance of normative conflict – but in a more subtle way, namely by refusing to allow generic norms to be construed or applied in isolation from more specialized rules. This makes sense because, as Grotius observed, in determining the true intentions of state parties, “... the preference is to be given to such [treaties] as are more particular, and approach nearer to the point in question.” But the importance of interpreting general rules in harmony with more specific rules does not advance McAdam’s thesis that the absence of rules to define the status of the broader class of non-returnable persons must be filled by the effective recasting of the Refugee Convention’s beneficiary class.
In sum, lex specialis is a general principle of law concerned with determining the relationship between norms. It exists primarily in order to resolve a conflict between two binding standards – not, as McAdam tacitly suggests, to fill a normative void. Because there is no normative conflict of either the direct or indirect variety – non-removable non-refugees can readily benefit from generic human rights without any infringement of the Refugee Convention’s special provisions for refugees – lex specialis has no legal relevance to the definition of the scope of the duty to protect non-refugees.
This is not to suggest that there are not good reasons in principle to extend many, if not all, Refugee Convention rights to the broader class of persons protected against refoulement. To the contrary, as observed above, the duty of equal protection may compel that result in at least some cases. But there is no extant legal basis to assert that all legally non-returnable persons are entitled de jure to claim all Refugee Convention rights.
Conceptual clarity on these issues matters. The net result of the persistent overstatement of the reach of international refugee law is not, as presumably hoped, the effective incorporation of new standards into a clear and practical system of enforceable duties. Consider, for example, the reaction of the English Court of Appeal when invited by UNHCR to find that the duty of non-refoulement should be deemed to have evolved beyond the text of Art. 33 in order to prohibit efforts to stymie the departure from their own countries of would-be refugees. UNHCR frankly acknowledged that its submissions to this end did “... not turn on the text of the Refugee Convention. Rather, they turn on understanding the international protection regime as a complex of international practice and precepts drawn from refugee law, human rights law, and general principles of international law... Where, as in the present case, issues arise that strictly do not fall within the Convention’s textual scope, its objectives and purposes should act as a reliable guide.”
The Court appropriately rejected this argument in clear terms. It cited with approval the view of the ICJ that “... although the principle of good faith is ‘one of the most basic principles concerned with the creation and performance of legal obligations... it is not in itself a source of obligation where none would otherwise exist’...” Most fundamentally, the Court refused to expand the duty of non-refoulement beyond what its text would reasonably bear simply because such an expansion would prove beneficial to at-risk persons. Adopting the earlier view of the High Court of Australia, the court asserted instead that
... the Convention, like many international and municipal instruments, does not necessarily pursue its primary purpose at all costs. The purpose of an instrument may instead be pursued in a limited way, reflecting the accommodation of the differing viewpoints, the desire for limited achievement of objectives, or the constraints imposed by limited resources... It would therefore be wrong to depart from the demands of language and context by invoking the humanitarian objectives of the Convention without appreciating the limits which the Convention itself places on the achievement of them.
It is of course true that the reform of international law, in particular
bodies of law such as refugee law which present few clear
strategic, much less
immediate, advantages to state parties, is a slow and often frustrating process.
But as the history of the last
two decades makes clear, it is not a Sisyphean
pursuit. We can already draw on an extraordinary expansion of the duty of
non-refoulement under the Civil and Political Covenant, as well as under
regional norms; we can build on the potential for major gains in this regard
under both the Convention on the Rights of the Child, and by reliance on
international humanitarian law; and we can ground the arguments
rights-attribution to non-refugee beneficiaries of protection against
refoulement in the duty of
While less glamorous and surely less
patient and incrementalist strategy allows us to pursue reform from within the
relatively secure space of legal obligation.
If in contrast the scope of extant
legal obligation is exaggerated, we impliedly jettison accrued gains and descend
into the realm
of pure policy – a space in which refugee rights are far
too often deemed dispensable in the pursuit of narrow definitions
In what may seem an ironic twist, those committed to expansion of the scope of protection must therefore concede that contrary to the claim of Lauterpacht and Bethlehem, there is no customary international legal obligation enjoining states not bound by relevant conventions to honor the duty of non-refoulement in relation to refugees and others facing the prospect of serious harm. And contrary to McAdam’s view, it is not the case that all persons entitled to protection against refoulement must, by virtue of a conceptual fusion of the Refugee Convention and other human rights accords, be granted all of the refugee-specific entitlements codified in the Refugee Convention itself.
There is, in short, no leveraged right to asylum.
[*] Dean of the
Melbourne Law School and William Hearn Professor of Law, University of
Melbourne. The research assistance of Anne Kallies,
as well as the thoughtful
comments of Michelle Foster, Martin Jones, Paul McDonough, Jason Pobjoy and
Michael Timmins on an earlier
draft are acknowledged with appreciation. This
analysis was initially presented at the Auckland University Faculty of Law
Rights at the Frontier’ Conference, Sept. 12, 2008. The
author thanks the New Zealand Legal Research Foundation and University
Auckland Law Faculty for the invitation to participate in this
 R. (Ullah) v. Special Adjudicator; Do v. Secretary of State for the Home Department,  UKHL 26, June 17, 2004. See generally N. Mole, Asylum and the European Convention on Human Rights (2007); and P. van Dijk et al eds., Theory and Practice of the European Convention on Human Rights (2006), at 433-440.
 See generally P. Alston and J. Crawford eds., The Future of UN Human Rights Treaty Monitoring (2000).
 UN Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, 1465 UNTS 85 (“Torture Convention”).
 International Covenant on Civil and Political Rights, 999 UNTS 172 (“Civil and Political Covenant”).
 See J. Hathaway, The Rights of Refugees under International Law (2005) (“Rights of Refugees”), at 368-370.
 1577 UNTS 3. See J. McAdam, Complementary Protection in International Refugee Law (2007) (“Complementary Protection”), at 173-196. It is of course true that relevant norms under these accords are more substantively circumscribed than the Refugee Convention’s open-ended focus on persons at risk of “being persecuted.” But protection claims derived from the Torture Convention, Civil and Political Covenant, Covention on the Rights of the Child and international humanitarian law are in other ways less constrained than is refugee status. In particular, there is no need to show the element of civil or political disfranchisement inherent in the Refugee Convention’s “for reasons of” nexus requirement; and states are not authorized to deny protection on the basis of lack of deservingness as Art. 1(F) of the Refugee Convention requires.
 See generally Hathaway, Rights of Refugees, at 369, fn.399.
 See generally Hathaway, Rights of Refugees, at 278-990.
 J. Hathaway, “The Evolution of Refugee Status in International Law: 1920-1950,” (1984) 33 ICLQ 348.
 Council Directive 2004/83/EC on minimum standards for the qualification and status of third country nationals or stateless persons as refugees or as persons who otherwise need international protection and the content of the protection granted, O.J. 304, 30/09/2004 P.0012-0023 (“Qualification Directive”).
 The original draft of the Qualification Directive (Proposal for a Council Directive on minimum standards for the qualification and status of third country nationals and stateless persons as refugees or as persons who otherwise need international protection”, EU Doc. 14643/02 (Asile 68), 22 November 2002), denied only two Refugee Convention rights to the beneficiaries of subsidiary protection. First, it was proposed that they would only be granted travel documents in limited circumstances. Second, access to the labor market was to be denied for up to six months following the date of status recognition: see J. Hathaway, “What’s in a Label?,” (2003) 5 European Journal of Migration and Law 1 (“Label”). The rights afforded to beneficiaries of subsidiary protection were, however, diluted as a consequence of further drafting. The limitation on access to travel documents was retained (Article 25). The six month limitation for access to the labor market was removed, but a discretion was introduced allowing Member States to take into account the situation of the labor market when considering potential limitations on access to the labor market for beneficiaries of subsidiary protection (Article 26). Discretionary exceptions were also introduced in relation to access to social welfare (Article 28), access to health care (Article 29) and the availability of family unification (Article 24). There is nonetheless a high correlation between the standards adopted for beneficiaries of subsidiary protection and Refugee Convention rights: Qualification Directive.
 The Commission of the European Communities has announced its intention to “reconsider the level of rights and benefits to be secured for beneficiaries of subsidiary protection, in order to enhance their access to social and economic entitlements which are crucial for their successful integration, whilst ensuring respect for the principle of family unity across the EU”: Commission of the European Communities, Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of Regions, “Policy Plan on Asylum: An Integrated Approach to Protection Across the EU,” COM (2008) 360 final (17 June 2008). The Commission’s proposal finds support in a Staff Working document accompanying the Communication: “Given the fact that in practical terms the situation of the two groups is comparable, their level of rights should also be (close to) equivalent. A clear example is the lack of provisions in EU law on family reunification for subsidiary protection beneficiaries. A higher level of rights for these persons is necessary if the EU wants to avoid creating a subclass of protected persons and also to respond to the call of the Hague programme which mention the established of a uniform protection status in the EU”: Commission of the European Communities, Commission Staff Working Document accompanying the Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of Regions, “Policy Plan on Asylum: An Integrated Approach to Protection Across the EU,” Impact Assessment, SEC (2008) 2029 (June 17, 2008).
 Immigration and Refugee Protection Act 2001, S.C. 2001, c. 27, at s. 97.
 For Australia, see Department of Immigration and Citizenship, “Draft Complementary Protection Model: Australia” (13 November 2008) (copy with author); United Nations High Commissioner for Refugees, “Draft Complementary Protection Visa Model: Australia: UNHCR Comments” (January 2009). The proposed model evidences an intention to provide an identical status to refugees and to beneficiaries of complementary protection. A similar approach has been proposed in New Zealand. The most recent draft of the Immigration Bill 2007 (NZ) is available at http://www.legislation.govt.nz/bill/government/2007/0132-2/latest/versions.aspx (accessed Aug. 7, 2009). See generally J. Pobjoy, “Treating like alike: The principle of non-discrimination as a tool to mandate the equal treatment of involuntary aliens entitled to international protection” (“Treating like alike”), (forthcoming, copy with author).
 A right to asylum was rejected both in the drafting of the Refugee Convention, and at the 1977 Territorial Asylum Conference. See generally A. Grahl-Madsen, Territorial Asylum (1980). I summarize the contemporary significance of the abortive effort to draft a binding commitment to grant asylum in The Law of Refugee Status (1991), at 13-16.
 See eg. J. Hathaway, “Why Refugee Law Still Matters,” (2007) 8 Melbourne Journal of International Law 89.
 E. Lauterpacht and D. Bethlehem, “The Scope and Content of the Principle of Non-Refoulement,” in E. Feller ed., Refugee Protection in International Law 87 (2003) (“Non-Refoulement”), at para. 253. The exception is for persons who face a threat to “life, physical security, or liberty” not rising to the level of a risk of “torture or cruel, inhuman or degrading treatment or punishment,” whose entitlement to protection against refoulement can be trumped by “overriding reasons of national security or public safety”: id. at para. 253(c).
 North Sea Continental Shelf Cases (Federal Republic of Germany/Denmark; Federal Republic of Germany/Netherlands),  ICJ Rep. 3, cited in Lauterpacht and Bethlehem, “Non-Refoulement,” at para. 198.
 Lauterpacht and Bethlehem, “Non-Refoulement,” at para. 198.
 Lauterpacht and Bethlehem, “Non-Refoulement,” at paras. 201-208. They add for good measure, that there is an “evident lack of expressed objection by any state to the normative character of the principle of non-refoulement”: id. at para. 216.
 Lauterpacht and Bethlehem, “Non-Refoulement,” at paras. 209-210.
 Lauterpacht and Bethlehem, “Non-Refoulement,” at paras. 211-215.
 Lauterpacht and Bethlehem, “Non-Refoulement,” at para. 216. Somewhat confusingly, they also seem to suggest that non-refoulement is a general principle of international law, though they provide no analysis in support of that view: id.
 Asylum Case (Colombia v. Peru),  ICJ 266.
 Asylum Case (Colombia v. Peru),  ICJ 266, at 277.
 “It is of course axiomatic that the material of customary international law is to be looked for primarily in the actual practice and opinio juris of States, even though multilateral conventions may have an important role to play in recording and defining rules deriving from custom, or indeed in developing them”: Continental Shelf (Libyan Arab Jamahiriya v. Malta) Judgment,  ICJ Rep. 13, at para. 27. See also Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 183; and Legality of the Threat or Use of Nuclear Weapons,  ICJ Rep. 226, at para. 64.
 Thus, the norm must “... be of a fundamentally norm-creating character such as could be regarded as forming the basis of a general rule of law”: North Sea Continental Shelf Cases (Federal Republic of Germany/Denmark; Federal Republic of Germany/Netherlands),  ICJ Rep. 3, at para. 72. While there was some concern in the case that the “equidistance principle” invoked as treaty-based custom met this test given its secondary character (see id.), there is little doubt that the duty of non-refoulement is of a norm-creating character.
 North Sea Continental Shelf Cases (Federal Republic of Germany/Denmark; Federal Republic of Germany/Netherlands),  ICJ Rep. 3, at paras. 209-215. “... [T]he substance of [customary] law must be ‘looked for primarily in the actual practice and opinio juris of States’”: Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 64. Accord International Law Association (“ILA”), “Statement of Principles Applicable to the Formation of General Customary International Law” (2000) (“Customary International Law”), at para. 9.
 “At best, the recommendation made by the Council constitutes the point of departure of an administrative practice...”: Reservations to the Convention on the Prevention and Punishment of the Crime of Genocide,  ICJ Rep. 15, at 25.
 To be fair, customary international law is notoriously murky terrain. As Goldsmith and Posner write, “It is unclear which state acts count as evidence of a custom, or how broad consistent state practice must be to satisfy the custom requirement. It is also unclear what it means for a nation to follow a custom from a sense of legal obligation, or how one determines whether such an obligation exists”: J. Goldsmith and E. Posner, “A Theory of Customary International Law,” (1999) 66 U. Chi. L.R. 1113, at 1114. In the result, “international law arguments based on custom always suffer from a considerable degree of arbitrariness”: N. Petersen, “Customary Law Without Custom? Rules, Principles, and the Role of State Practice in International Norm Creation,” (2007-2008) 23 Am. U. Intl. L.R. 275, at 277.
 Anthony D’Amato has strongly criticized the ICJ for commencing with analysis of opinio juris (rather than with analysis of whether there is consistent relevant state practice) in the Nicaragua case: A. D’Amato, “Trashing Customary International Law,” (1987) 81 AJIL 101 (“Trashing”), at 102. But as Oscar Schacter has observed, “[e]ven if the [reversal] seemed to place the cart before the horse, it did not depart in principle from the basic postulate that binding custom was the result of the two elements: State practice and opinio juris”: O. Schachter, “New Custom: Power, Opinio Juris and Contrary Practice,” in J. Makarczyk ed., Theory of international law at the threshold of the 21st century: Essays in honour of Krzysztof Skubiszewski 531 (1996) (“New Custom”), at 534.
 Only if relevant state actions are “based on their being conscious of having a duty to [act in a particular way] would it be possible to speak of an international custom”: The Case of the SS ‘Lotus’, PCIJ Ser.A, no.10, at 28.
 K. Wolfke, Custom in Present International Law (1993) (“Custom”), at 51. See also International Law Association, “Statement of Principles Applicable to the Formation of General Customary International Law” (2000), at 10: “...[T]he main function of the subjective elements is to indicate what practice counts (or, more precisely, does not count) towards the formation of a customary rule.” As Kammerhofer writes, “[t]he concept of opinio juris is arguably the centrepiece of customary international law. It is the most disputed, least comprehended component of the workings of customary international law. At the heart of the debate lies an important conflict: on the one hand, customary law-making seems by nature indirect and unintentional. On the other hand, law-making normally requires some form of intentional activity, an act of will. In the international legal system, great value has traditionally been placed in the states’ agreement or consent to create legal obligations binding on them...”: J. Kammerhofer, “Uncertainty in the Formal Sources of International Law: Customary International Law and Some of Its Problems,” (2004) 15 EJIL 523 (“Uncertainty”), at 532.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at paras. 188-190.
 See eg. R. Jennings and A. Watts eds., Oppenheim’s International Law (2008) (“Oppenheim’s”), at 28; I. Brownlie, Principles of Public International Law (2008) (“Principles”), at 6. But see J. Kelly, “The Twilight of Customary International Law,” (2000) 40 Va. J. Intl. L. 449 (“Twilight”), at 487: “Aspirational or recommendatory instruments, enacted while states remain unwilling to sign concrete treaties, provide compelling evidence that states lack the normative conviction necessary to create customary obligations, rather than evidence that states believe these norms are binding.”
 Lauterpacht and Bethlehem, “Non-Refoulement,” at para. 209.
 Lauterpacht and Bethlehem, “Non-Refoulement,” at para. 209.
 Writing in relation to the practice component of customary law, Villiger observes that “the condition of uniform practice requires that the instances of practice of individual States and of States in general circumscribe, apply, or refer to, and thereby express, the same customary rule”: M. Villiger, Customary International Law and Treaties: A Manual of Theory and Practice of the Interrelation of Sources (1997) (“Interrelation of Sources”), at 43.
 Lauterpacht and Bethlehem, “Non-Refoulement,” at para. 210.
 Lauterpacht and Bethlehem, “Non-Refoulement,” at para. 216.
 UNGA Res. 2312 (XXII), adopted Dec. 14, 1967.
 Id. at Art. 1(1) and Preamble.
 Specifically, persons threatened with persecution are one of the three groups said by Lauterpacht and Bethlehem to be entitled to protection against refoulement under a general customary duty. The other two are persons who face “a real risk of torture or cruel, inhuman or degrading treatment or punishment” and persons who face “a threat to life, physical integrity, or liberty”: Lauterpacht and Bethlehem, “Non-Refoulement,” at para. 218.
 There is a not-insignificant policy concern, noted by Thomas Franck. “The effect of [an] enlarged concept of the lawmaking force of General Assembly resolutions may well be to caution states to vote against ‘aspirational’ instruments if they do not intend to embrace them totally and at once, regardless of circumstances. That would be unfortunate. Aspirational resolutions have long occupied, however uncomfortably, a twilight zone between ‘hard’ treaty law and the normative void. Even if passed with a degree of cynicism, they may still have a bearing on the direction of normative evolution. By seeking to harden this ‘soft’ law prematurely, however, the [ICJ] advises prudent states to vote against such resolutions, or at least to abstain”: T. Franck, “Some Observations on the ICJ’s Procedural and Substantive Innovations,” (1987) 81 AJIL 116 (“Innovations”), at 119.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 188.
 Some commentators take strong objection to this holding. “... [A] customary rule arises out of state practice; it is not necessarily to be found in UN resolutions and other majoritarian political documents... If voting for a UN resolution means investing it with opinio juris, then the latter has no independent content: one may simply apply the UN resolution as it is and mislabel it ‘customary law’”: D’Amato, Trashing, at 102. This critique is overstated, as the ICJ merely held that General Assembly resolutions could contribute to opinio juris; consistent state practice must also be identified. D’Amato no doubt makes his charge in view of the Court’s regrettable assumption (rather than interrogation) of consistent state practice. The judgment is, however, clear that consistent state practice remains an essential element of customary international law formation: Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 184. Accord ILA, Customary International Law, at 63: “Given that General Assembly resolutions are not, in principle, binding, something more is needed to establish [opinio juris] than a mere affirmative vote (or failure to oppose a resolution adopted by consensus).”
 Legality of the Threat or Use of Nuclear Weapons,  ICJ Rep. 226, at para. 71. “... [I]t is necessary to look at its content and the conditions of its adoption; it is also necessary to see whether an opinio juris exists as to its normative character. Or a series of resolutions may show the gradual evolution of the opinio juris required for the establishment of a new rule”: id. An extreme interpretation is that “[t]his decision goes much farther than its predecessors in transforming [General Assembly resolutions] from exhortations or ‘soft law’ principles into ‘hard law’ prescriptions, at least in the eyes of the Court... Every resolution that purports to express a legal norm, even a ‘soft law’ exhortation or aspiration, has the potential of being recognized by the Court as a binding and strictly enforceable obligation, at least for those states which did not expressly dissent from it”: F. Morrison, “Legal Issues in the Nicaragua Opinion,” (1987) 81 AJIL 160, at 161. As James Crawford helpfully reminds us, “[o]f course, the General Assembly is not a legislature. Mostly its resolutions are only recommendations, and it has no capacity to impose new legal obligations on states”: J. Crawford, The Creation of States in International Law (2006) (“Creation of States”), at 113.
 See note 15 supra.
 E. Lapenna, “Territorial Asylum - Developments from 1961 to 1977 - Comments on the Conference of Plenipotentiaries,” (1978) 16 A.W.R. Bull. 1, at 4.
 A helpful contrast is provided by the facts of the Fisheries Jurisdiction Case (United Kingdom v. Iceland),  ICJ Rep. 3, at para. 56, noting that the opinio juris contended for by Iceland – a provision for special treatment of states overwhelmingly dependent on fishing – initially “failed to obtain the majority required, but a resolution was adopted at the 1958 Conference concerning the situation of countries or territories whose people are overwhelmingly dependent upon coastal fisheries for their livelihood or economic development.”
 This argument is, however, made by the UNHCR. “The principle of non-refoulement has been consistently referred to by the United Nations General Assembly in its various resolutions on the High Commissioner’s Annual Report. The Office of UNHCR considers that these references to the principle of non-refoulement, taken together with the... Conclusions of the [UNHCR] Executive Committee constitute further evidence of its acceptance as a basic normative principle”: UNHCR, “The Principle of Non-Refoulement as a Norm of Customary International Law: Response to the Questions Posed to UNHCR by the Federal Constitutional Court of the Federal Republic of Germany in Cases 2 BvR 1938/93, 2 BvR 1953/93, 2 BvR 1954/93,” Jan. 31, 1994, at para. 43.
 See eg. the references to non-refoulement in resolutions adopted routinely by the General Assembly upon receiving the High Commissioner’s annual report: UNGA Res. 38/121, adopted Dec. 16, 1983; UNGA Res. 39/140, adopted Dec. 14, 1984; UNGA Res. 40/118, adopted Decd. 13, 1985; UNGA Res. 41/124, adopted Dec. 4, 1986; UNGA Res. 42/109, adopted Dec. 7, 1987; UNGA Res. 43/117, adopted Dec. 8, 1988; UNGA Res. 44/137, adopted Dec. 15, 1989; UNGA Res. 45/140, adopted Dec. 14, 1990; UNGA Res. 46/106, adopted Dec. 16, 1991; UNGA Res. 47/105, adopted Dec. 16, 1992; UNGA Res. 48/116, adopted Dec. 20, 1993; UNGA Res. 49/169, adopted Dec. 23, 1994; UNGA Res. 50/152, adopted Dec. 21, 1995; UNGA Res. 51/75, adopted Dec. 12, 1996; UNGA Res. 52/103, adopted Dec. 12, 1997; UNGA Res. 53/125, adopted Dec. 9, 1998; UNGA Res. 54/146, adopted Dec. 17, 1999; UNGA Res. 55/74, adopted Dec. 4, 2000; UNGA Res. 56/137, adopted Dec. 19, 2001; UNGA Res. 57/187, adopted Dec. 18, 2002; UNGA Res. 58/151, adopted Dec. 22, 2003; UNGA Res. 59/170, adopted Dec. 20, 2004; UNGA Res. 60/129, adopted Dec. 16, 2005; UNGA Res. 61/137, adopted Dec. 19, 2006; UNGA Res. 62/124, adopted Dec. 18, 2007; UNGA Res. 63/148, adopted Dec. 18, 2008.
 See text at note 17 supra.
 There is no pattern of substantial negative votes or abstentions of a kind that would negate the opinio juris value of the resolutions: Legality of the Threat or Use of Nuclear Weapons,  ICJ Rep. 226, at para. 71.
 “To begin with, over half the States concerned, whether acting unilaterally or conjointly, were or shortly [afterward] became parties to the Geneva Convention, and were therefore presumably, so far as they were concerned, acting actually or potentially in the application of the Convention. From their action no inference could legitimately be drawn as to the existence of a rule of customary international law in favour of the equidistance principle”: North Sea Continental Shelf Cases (Federal Republic of Germany/Denmark; Federal Republic of Germany/Netherlands),  ICJ Rep. 3, at para. 76.
 For example, the Thai Ministry of the Interior lists as one of its key functions the effort “to intercept and drive back refugees”: U.S. for Refugees and Immigrants (USCRI), World Refugee Survey 2008 (2008), at 7. The Malaysian Minister of Information similarly announced with respect to Acehnese refugees, “We will treat them as we do other refugees. We will detain them and send them back”: Reuters, “Malaysia says it will deport fleeing Acehnese,” June 3, 2003, quoting Khalil Yaacoub. In India, there is no domestic legal framework for recognizing refugees. The Foreigners Act 1946 does not distinguish between undocumented migrants and refugees, and allows the government to arrest, detain and deport any undocumented migrant. See generally P. Saxena, “Creating Legal Space for Refugees in India: the Milestones crossed and the Roadmap for the Future,” (2007) 19(2) IJRL 246.
 Reliance is sometimes placed on express acknowledgments of the duty of non-refoulement in bilateral arrangements between regional states and the UNHCR, but these are not in fact a dependable indicator of opinio juris. For example, despite having executed such an agreement Jordan simply closed its borders to Palestinian and Iranian Kurdish refugees in 2006 on the basis of capacity and concerns that the refugees would not depart even when the risk abated: Human Rights Watch, “Nowhere to Flee: The Perilous Situation of Palestinians in Iraq,” Sept. 2006 (“Nowhere to Flee”), at 38.
 Participation in both the Torture Convention, supra note 3 (containing an express duty of non-refoulement in Art. 3) and the Civil and Political Covenant, supra note 4 (containing an implied duty of non-refoulement in relation to Arts. 6 and 7) is fairly strong. 70% of Asian states and 85% of Middle Eastern countries are parties to the Torture Convention, while 80% of Asian countries and 85% of Middle Eastern nations are parties to the Civil and Political Covenant: Office of the High Commissioner on Human Rights, http://www2.ohchr.org/english/bodies/ratification (as at 14 August 2009). Notwithstanding the high participation rate, only two Asian states (Kazakhstan and the Republic of Korea) and one Middle Eastern country (Qatar) have accepted the competence of the Committee Against Torture in accordance with Article 22 of the Torture Convention. Similarly, only 50% of Asian state parties to the Civil and Political Covenant and no Middle Eastern country have accepted the competence of the Human Rights Committee to hear individual complaints under the Optional Protocol to the Civil and Political Covenant (99 UNTS 302): Office of the High Commissioner for Human Rights, http://www2.ohchr.org/english/bodies/ratification (as at 15 August 2009). Acceptance of a duty of non-refoulement vis à vis refugees in both of these regions is very low. Only 40% of Asian countries and 23% of Middle Eastern states have acceded to either the Refugee Convention or Protocol: UNHCR, “State Parties to the 1951 Convention relating to the Status of Refugees and the 1967 Protocol,” (Oct. 1, 2008). This refusal formally to be bound by the duty to avoid the refoulement of refugees is moreover of long-standing: see K. Hailbronner, "Nonrefoulement and 'Humanitarian' Refugees: Customary International Law or Wishful Legal Thinking?", in D. Martin ed., The New Asylum Seekers: Refugee Law in the 1980s (1988), at 128-129. The claim of opinio juris in support of a comprehensive duty of non-refoulement is thus undermined.
 Lebanon’s Memorandum of Understanding with the UNHCR, for example, states that “Lebanon does not consider itself an asylum country”: United Nations High Commissioner for Refugees, “UNHCR Global Report 2003 – The Middle East Regional Overview,” (June 1, 2004). Lebanon therefore permits refugees to remain only on the condition that they are resettled or repatriated by the UNHCR within a period of six months: USCRI, World Refugee Survey 2009 (online version, 2009): <http://www.refugees.org/countryreports.aspx?id=2334>. A similar concern arises in Jordan: see note 57 supra. See generally M. Kagan, “The Beleagured Gatekeeper: Protection Challenges Posed by UNHCR Refugee Status Determination,” (2006) 18(2) Int. J. Ref. L. 1.
 J.S. Verma, “Inaugural Address,” in UNHCR and SAARCLAW eds., Seminar Report: Refugees in the SAARC Region: Building a Legal Framework (1997), at 13-18. Accord P. Saxena, “Creating Legal Space for Refugees in India: the Milestones Crossed and the Roadmap for the Future,” (2007) 19(2) Intl. J. Ref. L. 246, at 255: “A plethora of unreported cases demonstrates that the courts have treated these matters on purely technical grounds; no pronouncements of law are made nor are any general guidelines laid. This explains why the majority of these cases do not find a place in law reports. Interim non-speaking orders may provide relief in individual cases, but their contribution to jurisprudence is negligible, even negative at times. Ranabir Samaddar has agreed that the judicial reasoning has been mainly humanitarian and not rights based, dispensing kindness and not justice, and that the Court has nothing to say on the ‘refugee-situation.’” See also O. Chaudhary, “Turning Back: An Assessment of Non-Refoulement under Indian Law,” (2004) 39 Economic and Political Weekly 3257. But see V. Vijayakumar, “Judicial Responses to Refugee Protection in India,” (2000) 12(2) Intl. J. Ref. L. 235, at 235-236, arguing that Indian court decisions have provided “a series of rights to the millions of refugees who had to cross the internationally recognized borders and continue to stay in India.”
 “As regards those States, on the other hand, which were not, and have not become parties to the Convention, the basis of their action can only be problematical and must remain entirely speculative. Clearly, they were not applying the Convention. But from that no inference could justifiably be drawn that they believed themselves to be applying a mandatory rule of customary international law. There is not a shred of evidence that they did and... there is no lack of other reasons for using the equidistance method, so that acting, or agreeing to act in a certain way, does not of itself demonstrate anything of a juridical nature... The frequency, or even habitual character of the acts is not itself enough”: North Sea Continental Shelf Cases (Federal Republic of Germany/Denmark; Federal Republic of Germany/Netherlands),  ICJ Rep. 3, at paras. 76-77.
 “... State practice, including that of States whose interests are specially affected, should have been both extensive and virtually uniform in the sense of the provision invoked; and should moreover have occurred in such a way as to show a general recognition that a rule of law or legal obligations is involved”: North Sea Continental Shelf Cases (Federal Republic of Germany/Denmark; Federal Republic of Germany/Netherlands),  ICJ Rep. 3, at para. 74.
 At the end of 2008, 6,228,400 refugees were in the Middle East; 909,100 in East Asia and the Pacific; and 2,512,400 were in South and Central Asia. This amounts to more than 70% of the total world refugee population: USCRI, World Refugee Survey 2009 (2009), at 33.
 Hudson’s classic definition speaks of four elements, including “... (a) concordant practice by a number of States with reference to a type of situation falling within the domain of international relations; (b) continuation or repetition of the practice over a considerable period of time; (c) conception that the practice is required by, or consistent with, prevailing international law; and (d) general acquiescence in the practice by other States”: M. Hudson,  2 YBILC 26, U.N. Doc. A/CN.4/SER.A/1950/Add.1. Elements (a), (b), and (d) have converged over time in the requirement to demonstrate that “... the conduct of States should, in general, be consistent with [the putative norm]...: Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 186. Yet “[i]t is not to be expected that in the practice of States the application of the rules in question should have been perfect, in the sense that States should have refrained, with complete consistency, from [actions prohibited by the putative norm]”: id. at para. 185. Hudson’s element (c) remains a second and independent criterion for recognition of a rule of customary international law. “It is of course axiomatic that the material of customary international law is to be looked for primarily in the actual practice and opinio juris of States...”: Continental Shelf (Libyan Arab Jamahiriya v. Malta) Judgment,  ICJ Rep. 13, at para.27.
 USCRI, World Refugee Survey 2008 (2008), at 22.
 The most basic categories of Refugee Convention rights inhere provisionally in persons claiming to be refugees until and unless they are determined not to qualify for refugee status: Hathaway, Rights of Refugees, at 156-160.
 Hathaway, Rights of Refugees, at 279-281.
 Id. at 281-282.
 Id. at 282.
 Id. at 283-284.
 Id. at 284-286.
 Id. at 287.
 Id. at 287-289.
 Id. at 290-291.
 Id. at 291-299.
 Indeed, the United Nations Commission on Human Rights has formally expressed its “distress” at the “widespread violation of the principle of non-refoulement and of the rights of refugees”: UN Commission on Human Rights, Res. 1997/75.
 Asylum Case (Colombia v. Peru),  ICJ 266, at 276.
 North Sea Continental Shelf Cases (Federal Republic of Germany/Denmark; Federal Republic of Germany/Netherlands),  ICJ Rep. 3, at para. 74.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 186.
 Brownlie, Principles, at 7. Hersch Lauterpacht cautions, however, that “because of the underlying requirement of consent, the condition of constancy and uniformity is liable on occasion to be interpreted with some rigidity when there is a question of ascertaining a customary rule of general validity”: E. Lauterpacht ed., International Law: The Collected Papers of Hersch Lauterpacht 62 (1970).
 Asylum Case (Colombia v. Peru),  ICJ 266, at 277.
 “State practice, the material element, provides the concrete evidence of normative conviction”: Kelly, Twilight, at 500.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 202.
 In the same paragraph, the Court found that “[t]he existence in the opinio juris of States of the principle of non-intervention is backed by established and substantial practice”: Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 202.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 207.
 As such, Duffy’s conclusion that the fact“[t]hat states have rarely totally disregarded their duty not to ‘refoule’ individuals to face torture is evidence of the normative practice of non-refoulement (emphasis added)” is not justified: A. Duffy, “Expulsion to Face Torture? Non-refoulement in International Law,” (2008) 20(3) Intl. J. Ref. L. 373, at 387.
 Villiger, Interrelation of Sources, at 29.
 See text supra at note 65.
 See text supra at note 66.
 See text supra at note 65. This is not to say that states outside of Asia and the Middle East have solid records of avoiding refoulement. To the contrary, states around the world have often violated the putative norm: see text supra at notes 67-76.
 Of the 24 states assessed as presenting “systemic” or “severe”risks of refoulement, half are in Asia or the Middle East, namely China, Egypt, Iran, Iraq, Israel (including the Occupied Territories), Lebanon, Libya, Malaysia, Pakistan, Saudi Arabia, Syria, and Yemen: USCRI, World Refugee Survey 2009 (2009), at 22.
 Villiger, Interrelation of Sources, at 44.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 186.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 186.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 206.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at paras. 207-208.
 For example, Egypt summarily sent Sudanese refugees back to Sudan in 2007, as well as Eritrean refugees to Eritrea in 2008 with no explanation or justification given: Amnesty International, “Egypt/Israel: Fear For Safety,” Doc. MED 15/038/2008/UA/241/08, Sept. 3, 2008. US President Bush simply declared, “We will turn back any refugees that attempts to reach our shore, and that message needs to be very clear as well to the Haitian people”: Human Rights Watch, “US: Don’t Turn Away Haitian Refugees,” Feb. 26, 2004. As Kelly observes, “[n]ations do not regularly explain the legal basis of their actions, nor is it clear how to determine the normative belief of hundreds of states, many of whom have never had the opportunity or need to express their opinion on a particular principle”: Kelly, Twilight, at 470.
 Greece has asserted that whole groups of persons seeking protection are not refugees, treating them simply as unauthorized migrants. “The Greek coast guard systematically forced boatloads of potential asylum seekers out of its national waters and back into Turkish territorial waters, sometimes deliberately damaging their boats to prevent their return or attempting to swamp them with waves, and, occasionally abandoning migrants on uninhabited islands...Greek border guards arrested migrants upon arrival, issued all of them automatic deportation orders, and detained them incommunicado without registration for several days before returning them to Turkey”: USCRI, World Refugee Survey 2008 (2008, online version), http://www.refugees.org/countryreports.aspx?id=2138 (accessed 15 August 2009). See also S. Troller, “Greece does EU’s migration dirty work,” The Guardian (Jan. 25, 2009), <http://www.guardian.co.uk/commentisfree/2009/jan/25/eu-greece>. Jordan insisted that it had the right to refuse entry to Iraqi Palestinians on the grounds of the enormity of its other responsibilities towards the Palestinians: Human Rights Watch, Nowhere to Flee, at 38.
 Lauterpacht and Bethlehem argue that the only internal limitation to the putative customary norm is where a state demonstrates “[o]verriding reasons of national security or safety... in circumstances in which the threat of persecution does not equate to and would not be regarded as being on par with a danger of torture or cruel, inhuman or degrading treatment or punishment and would not come within the scope of other non-derogable customary principles of human rights. The application of these exceptions is conditional on the strict compliance with principles of due process of law and the requirements that all reasonable steps must first be taken to secure the admission of the individual concerned to a safe third country”: Lauterpacht and Bethlehem, Non-Refoulement, at para. 253.
 Examples of the classic opposition are those who assert that only physical acts count as practice, eg. A. D’Amato, The Concept of Custom in International Law (1971); and, arguing that custom may be based on verbal acts alone, B. Cheng, “Custom: The Future of General State Practice in a Divided World,” in R. Macdonald and D. Johnston eds., The Structure and Process of International Law: Essays in Legal Philosophy Doctrine and Theory 532 (1983).
 Kammerhofer, Uncertainty, at 525.
 See text supra at note 28.
 ILA, “Customary International Law.”
 ILA, “Customary International Law,” at para. 10(c).
 See eg. Franck, Innovations, at 118-119; S. Donaghue, “Normative Habits, Genuine Beliefs and Evolving Law: Nicaragua and the Theory of Customary International Law,” (1995) 16 AYBIL 327 (“Normative Habits”), at 338; Villiger, Interrelation of Sources, at 20.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 188.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 183, quoting from Continental Shelf (Libyan Arab Jamahiriya v. Malta) Judgment,  ICJ Rep. 13, at para. 27.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 184.
 “In Nicaragua... the ICJ discussed the requirement of state practice, but neither analyzed, nor cited examples of this element”: Kelly, Twilight, at 476, fn.112. See also Franck, Innovations, at 118-119, and F. Kirgis, “Custom on a Sliding Scale,” (1987) 81 AJIL 146 (“Sliding Scale”), at 147.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 186.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 186.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 186.
 Franck, Innovations, at 118-119; Kirgis, Sliding Scale, at 147.
 Having found there to be “abstention” from the use of force other than as authorized by the UN Charter, the Court turned to the issue of opinio juris: Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 188.
 See text supra at note 34.
 Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at paras. 188-190.
 “The Court has however to be satisfied that there exists in customary international law an opinio juris as to the binding character of such abstention. This opinio juris may, though with all due caution, be deduced from, inter alia, the attitude of the Parties and the attitude of States towards certainly General Assembly resolutions... It would therefore seem apparent that the attitude referred to expresses an opinio juris respecting such rules (or set of rules), to be thenceforth treated separately from the provisions, especially those of an institutional kind, to which it is subject on the treaty-law plane of the Charter”: Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at para. 188.
 The arguments made by Donaghue – that verbal “practice” is neither more ambiguous in purport nor necessarily any more politically motivated than practice on the ground – are not really arguments in favour of treating verbiage as practice; they are rather counterpoints to arguments made against this position: Donaghue, Normative Habits, at 332.
 Villiger, Interrelation of Sources, at 21.
 The primary meaning of “practice,” however, focuses on “habitual action or performance”; it is “action or execution as opposed to theory”: Concise Oxford Dictionary (1995), at 1072.
 See text supra at note 106. But see text at notes 110-117, indicating why a careful reading argues against this interpretation.
 See text supra at note 28.
 Villiger, Interrelation of Sources, at 22.
 Schachter, New Custom, at 533-534.
 Schachter, New Custom, at 534-535.
 “Perhaps the difficulty arises in part from the attempt to differentiate too clearly between practice and the opinio juris. They are rather aspects of the same idea. Even the older writers do not always mean by ‘practice’ the mere habit of acting in a certain way but rather the evidence, in the form of dispatches, opinions, arguments and so on, which support the existence of an opinio juris. Seen thus, the possible effect of a generally acclaimed General Assembly resolution falls easily into place in the orthodox scheme of things”: R. Jennings, Collected Writings of Sir Robert Jennings 10 (1998).
 Others are more cautious in this regard. “Collective resolutions by States in international organizations are not sufficient by themselves to generate customary norms. There has to be evidence of additional State practice which is consistent with those collective resolutions”: L. Hannikainen, Peremptory Norms (Jus Cogens) in International Law: Historical Development, Criteria, Present Status 236 (1988). See also Crawford, Creation of States, at 114: “State practice is just as much State practice when it occurs in the context of the General Assembly as in bilateral fora. The practice of States in assenting to and acting upon law-declaring resolutions may be of probative importance, in particular where that practice achieves reasonable consistency over a period of time.”
 Jennings and Watt, Oppenheim’s, at 31. Accord Villiger, Interrelation of Sources, at 21: “For most members of the State community, the UN and similar bodies have become the most important fora in which to express themselves collectively or individually.”
 Charter of the United Nations, 59 Stat. 1031 (1945), at Arts. 10-18. These articles authorize the General Assembly to make binding decisions only on a range of administrative matters.
 “... [S]tates could amend the UN Charter to create a new, more democratic process at the UN General Assembly. Similarly, resolutions passed in a prescribed form could bind all members specifically voting for a measure. States could approve an even more radical measure that would specifically bind all states to norms upon passage of a law-defining resolution by an appropriate supermajority”: Kelly, Twilight, at 497.
 Schachter, New Custom, at 534.
 Convention on the Prevention and Punishment of the Crime of Genocide, 78 UNTS 277; Rome Statute of the International Criminal Court, 2187 UNTS 90.
 A weaker version of this thesis is that treating words as practice“has many beneficial functions,” in particular ease of access to documentation of practice and the ability to change international law without breaking it: Villiger, Interrelation of Sources, at 21-22. While these technical points have merit, it is difficult to imagine that either is so pressing that it justifies a revision of a core rule of international lawmaking.
 Schachter, New Custom, at 538.
 Schachter, New Custom, at 538.
 Schachter, New Custom, at 538. He cites in particular “the prohibitions of aggression, genocide, slavery, torture and systematic racial discrimination... [and] the humanitarian law of armed conflict”: id. at 538-539. He distinguishes these areas from “the law on jurisdiction, immunities, State responsibility, [and] diplomatic privileges” where he does not believe the requirement of consistent state practice can be satisfied by words alone: id. at 538. Other commentators reach the same conclusion. See eg. J. Wouters and C. Ryngaert, “Impact on the Process of the Formation of Customary International Law,” in M. Kamminga and M. Scheinin eds., The Impact of Human Rights Law on General International Law 111 (2009) (“Impact”), at 112: “It will be argued that the more important the common interests of states or humanity are, the greater the weight that may be attached to opinio juris as opposed to state practice. If the stakes are high, inconsistent state practice may be glossed over, and a high premium may be put on states’ statements and declarations, inter alia in multilateral fora, in identifying customary law combined with general principles of law.”
 Schachter, New Custom, at 539.
 Kirgis, Sliding Scale, at 149. But as Donaghue observes, “[a]rguments that practice, or opinio juris... form part of a sliding scale are clearly incorrect as they fail to recognize the purpose of the inclusion of these elements in Article 38 [of the ICJ Statute]”: Donaghue, Normative Habits, at 330-331.
 Kirgis, Sliding Scale, at 149. See also ILA, “Customary International Law,” at 13: “When defining State practice – the objective element in customary law – it is necessary to take account of the distinction between what conduct counts as State practice and the weight to be given to it.”
 Kirgis, Sliding Scale, at 148-149. “Views questioning the necessity of one of the... two elements... have no foundation in international legal practice”: Wolfke, Custom, at 41.
 See text supra at notes 106-118.
 Kelly appropriately refers to these theories as “normative discourse masquerading as empirical”: Kelly, Twilight, at 497.
 Wouters and Ryngaert, Impact, at 112.
 “... [T]hat customary international law furnishes a means to develop universal norms when actual agreement is difficult or inconvenient cannot justify norms when there is no genuine acceptance”: Kelly, Twilight, at 537.
 “Without practice (consuetudo), customary international law would obviously be a misnomer, since practice constitutes precisely the main differentia specifica of that kind of international law”: Wolfke, Custom, at 40-41.
 “... Customary law is by its very nature the result of an informal process of rule-creation, so that the degree of precision found in more formal processes of law-making is not to be expected here”: ILA, “Customary International Law,” at 2.
 See text supra at notes 32-35.
 Kelly, Twilight, at 486.
 The resolutions of the General Assembly may provide evidence of opinio juris, or confirm the existence of a norm of customary international law: Legality of the Threat or Use of Nuclear Weapons,  ICJ Rep 226, at para. 70. It remains the case, however, that inconsistent state practice precludes the development of a customary norm despite strong evidence of opinio juris: id. at para. 73.
 The International Court of Justice has taken the position that “[w]hen it is the intention of the State making [a] declaration that it should become bound according to its terms, that intention confers on the declaration the character of a legal undertaking, the State being thenceforth legally required to follow a course of conduct consistent with that declaration. An undertaking of this kind, if given publicly, and with an intent to be bound, even though not made within the context of international negotiations, is binding”: Australia v. France,  ICJ Rep 253, at 267; New Zealand v. France,  ICJ Rep 457, at 472; endorsed in Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States, Merits),  ICJ Rep. 14, at paras. 39-40. It seems clearly to have been the Court’s intention to constrain this doctrine; however, the same result could readily have been avoided by reliance on such general principles of law as acquiescence or estoppel. A WTO panel has appropriately urged caution in the application of this approach, noting that “[a]ttributing international legal significance to unilateral statements made by a State should not be done lightly and should be subject to strict conditions...”: WTO Panel Report, United States – Sections 301-310 of the Trade Act of 1974, Dec. No. WT/DS152/R, Jan. 27, 2000, at para. 7.118.
 “The misunderstanding resulting from such a broad interpretation [of state practice] arises from the fact that it neglects the very essence of every kind of custom, which for centuries has been based upon material deeds and not words”: Wolfke, Custom, at 41-42.
 “The strategic advantage of elevating customary international law to a rule of recognition is that it allows the theorist to redefine the requirements of customary international law from empirical law to a preferred process while retaining its formal authority”: Kelly, Twilight, at 494.
 Wolfke, Custom, at 42.
 See Kelly, Twilight, at 495-496.
 “Normative scholars, advocates, and self-interested states are misusing an empirical source of law to articulate their preferred norms as if they were propounding a constitution rooted in common culture... I do believe that in a diverse world without a consensus on values, a general normative approach is premature and would threaten primary values, such as state sovereignty and the procedural values of open, democratic decision-making, that retain vitality”: Kelly, Twilight, at 458.
 “Approximately two centuries after the rise of the positivist view, a new theory [of customary international law] is beginning to take hold in some quarters. The theory derives norms of CIL in a loose way from treaties (ratified or not), UN General Assembly resolutions, international commissions, and academic commentary – but all colored by a moralism reminiscent of the natural law view”: J. Goldsmith and E. Posner, “Understanding the Resemblance Between Modern and Traditional Customary International Law,” (2000) 40(2) Virginia Journal of International Law 639, at 640.
 Kelly, Twilight, at 467.
 Schachter, New Custom, at 538.
 Wouters and Ryngaert, Impact, at 112.
 Kirgis, Sliding Scale, at 149.
 Kelly, Twilight, at 498.
 Kelly, Twilight, at 492.
 “Under the indeterminate and manipulable theory of customary international law... customary international law is then a matter of taste”: Kelly, Twilight, at 451.
 “... [R]esolutions and recommendations..., however sympathetic one may be towards their motivation and purpose, cannot themselves establish a legal rule binding in international law”: Sepet and Bulbul v. Secretary of State for the Home Department,  UKHL 15 (UK HL, Mar. 20, 2003). See also Garza v. Lappin  USCA7 303; (2001) 253 F 3d 918 (US CA7), at 924-925: “The American Declaration of the Rights and Duties of man, on which the Commission relied in reaching its conclusions in Garza’s case, is an aspirational document which, as Garza admitted in his petition... did not on its own create an enforceable obligation on the part of the OAS member nations.”
 Kelly, Twilight, at 495. “The substantive norms offered as customary international law in much of the Western literature are, not coincidentally, norms associated with individualism and the market economy”: id. at 466.
 Kelly, Twilight, at 489. As Byers has observed, “[t]he newly independent non-industrialized States found themselves in a legal system which had been developed primarily by relatively wealthy, militarily powerful States. They consequently sought to change the system. They used their numerical majorities to adopt resolutions and declarations which advanced their interests. They also asserted, in conjunction with a significant number of legal scholars (and perhaps with the International Court of Justice) that resolutions and declarations are instances of State practice which are potentially creative, or at least indicative, of rules of customary international law... Powerful States, for the most part, along with some scholars from powerful States, have resisted these developments. They have emphatically denied that resolutions and declarations can be State practice”: M. Byers, Custom, Power and the Power of Rules: International Relations and Customary International Law 41 (1999).
 “... [T]he customary international law process does not encourage compliance. With few effective means of enforcing norms, the international system relies on commitment and reciprocal self-interest for compliance. Nations that played no role in the formation of norms nor had their interests considered are unlikely to honor such norms”: Kelly, Twilight, at 540-541.
 In the context of the asserted duty of non-refoulement, see eg. Hathaway, Rights of Refugees, at 279-300. But the customary legal argument recently found favour before Justice Hartmann of the Hong Kong Court of First Instance in a decision presently on appeal: C. v. Director of Immigration, Hong Kong Court of First Instance Dec. No. 132/2006, Feb. 18, 2008. Reviewing not only the Lauterpacht and Bethlehem opinion but also relevant UNHCR Executive Committee conclusions and the full range of scholarly positions, the judge determined that “... [o]n balance... it must be recognized that the principle of non-refoulement [as] it applies to refugees has grown beyond the confines of the Refugee Convention and has matured into a universal norm of customary international law”: id. at para. 113. Extraordinarily, the judgment makes this finding against an express acknowledgment that “... a good many states have... by their actions been unambiguous in their repudiation of the norm as it has evolved in customary international law” (id. at para. 114); indeed, it observes that UNHCR proclamation of the non-derogable nature of the customary duty of non-refoulement “... was made by the Executive Committee... against the backdrop of ‘widespread violations of the principle of non-refoulement’”: id. at para. 118. In the end, however, the court refused the declaration sought by the applicants on the ground that Hong Kong has been a persistent objector to the norm – a conclusion reached not on the basis of real evidence of persistent objection, but rather on legally doubtful basis that in the context of “... the refusal to accede to the Refugee Convention, the refusal to enlarge the terms of the Immigration Ordinance, the making of specific reservations concerning immigration and the often-stated policy against asylum – Hong Kong’s refusal to pass legislation incorporating the rule is equivalent to passing legislation for the purpose of excluding it”: id. at para. 149. For a detailed description of this case, see O. Jones, “Customary Non-Refoulement of Refugees and Automatic Incorporation into the Common Law: A Hong Kong Perspective,” (2009) 58 ICLQ 443.
 “If nations have, in fact, accepted legal norms and possess the necessary normative conviction, then the vast majority of states should have little difficulty signing a treaty. Modern communications and transportation have simplified the logistics of international meetings, reducing treaty negotiations and international decisionmaking to a common occurrence”: Kelly, Twilight, at 538.
 McAdam, Complementary Protection.
 The one critical exception asserted by McAdam relates to persons who would fall afoul of the exclusion clauses of the Refugee Convention: id. at 223-242.
 McAdam’s analysis of the beneficiary class is not grounded in customary international law. She focuses instead on the various treaty-based regimes which expressly or by interpretation give rise to a duty of non-refoulement. See id. at 53-196. There is, however, no basis in principle to distinguish the rights of persons entitled to protection against refoulement under a treaty from those entitled to the same protection by virtue of customary international law.
 “In the human rights context, however, [non-refoulement] has been separated from these other rights to provide the trigger for protection without any corresponding legal status. The result is a protection gap”: id. at 201.
 Hathaway, Rights of Refugees, at 91-93.
 The clear exception is Art. 3 of the Torture Convention, which contains an explicit duty of non-refoulement yet does not define the rights of the beneficiary class. Expert analysis of this treaty provides no explanation for the omission: M. Nowak and E. McArthur, The United Nations Against Torture: A Commentary (2008).
 “The strong theoretical claims of human rights law unfortunately do not always sit comfortably with the realities of State practice”: McAdam, Complementary Protection, at 253.
 “There is not yet a consistent understanding of what the resultant legal status [of the beneficiaries of complementary protection] should entail, although this book advances the argument that a status identical to Convention status ought to apply”: id. at 3; “Though a number of States have traditionally respected these additional non-refoulement obligations, they have been reluctant to grant beneficiaries a formal legal status analogous to that enjoyed by Convention refugees”: id. at 5; “States have sought to distinguish them from refugees by granting them fewer rights and entitlements”: id. at 11; “In many cases, healthcare, employment, social security, and other rights which Convention refugees receive are denied. Accordingly, the extent of protection may be little more than non-refoulement through time”: id. at 17.
 Id. at 210.
 Id. at 11.
 Id. at 198.
 Id. at 220.
 “... [T]here is no legal justification for differentiating the rights of beneficiaries of international protection based on the source of the protection need”: id. at 197.
 Civil and Political Covenant. McAdam does not, however, explicitly reference this critical provision. But see J. Hathaway, Label, at 8; and especially Pobjoy, Treating like alike.
 See Hathaway, Rights of Refugees, at 123-147.
 “... [T]he Convention operates as a lex specialis for all persons in need of international protection – a specialized blueprint for legal status, rights, and obligations, irrespective of the legal source of the protection obligation (emphasis added)”: McAdam, Complementary Protection, at 17; “... [B]eneficiaries of complementary protection are entitled to the same legal status as Convention refugees... (emphasis added): id. at 197; “... [T]he extended scope of non-refoulement under international human rights and humanitarian law imposes a two-fold obligation on States: to refrain from removing persons to territories where they face a substantial threat of particular kinds of ill-treatment; and to provide such persons with a legal status equivalent to that of Convention refugees (emphasis added)”: id. at 252.
 Indeed, McAdam speaks of the beneficiaries of complementary protection as “... refugees who fall outside the framework of the major international treaties, the 1951 Refugee Convention and the 1967 Protocol”: id. at 1.
 Id. at 11.
 Id. at 209.
 “Non-refoulement is certainly the most fundamental principle of refugee law – indeed, its application to persons in need of international protection might be described as ‘qualifying’ or ‘constitutive’ of their status. The question, who are the beneficiaries of international protection?, is a converse way of asking, who is protected by the principle of non-refoulement?”: id. at 200.
 Hathaway, Rights of Refugees, at 302 ff. Indeed, the duty of non-refoulement did not appear in pre-1933 refugee conventions, having been added then as an afterthought to fill a perceived void in the scope of Art. 32's duty of non-expulsion: id.
 McAdam, Complementary Protection, at 209.
 Id. at 11.
 Id. at 1.
 It should, however, be acknowledged that generic international human rights law in some instances offers protections that exceed the scope of guarantees in the Refugee Convention itself. For example, rights to both physical security and to access the necessities of life, not codified in refugee law, can be established for refugees and other non-citizens by reliance on the Human Rights Covenants. See Hathaway, Rights of Refugees, at 439-514.
 McAdam at some points attempts an historical argument. “... [I]f historical definitions are considered, then persons who today ‘only’ fall within complementary protection would in some cases have been recognized as refugees under formal legal definitions, and persons who today fall within article 1(A)(2) of the Refugee Convention may in the past have been denied protection”: id. at 198. Assuming the point to be accurate, it nonetheless has no necessary present legal significance for purposes of interpreting the (present, Art. 1(A)(2)-based) refugee definition.
 In the European context of primary concern to McAdam, however, there is actually little value to be secured by assimilation of the rights of subsidiary protection beneficiaries to those of Convention refugees. As set out at note 11 supra, the main difference in rights allocation between these groups within Europe concerns the length of residence permits and availability of family reunification – neither being a subject regulated by the Refugee Convention. Non-discrimination analysis is thus a more plausible basis to close the gap on these and most other points. See Pobjoy, Treating like alike.
 “... [H]uman rights law alone does not provide a sufficient status for beneficiaries of complementary protection. Despite the theoretical universality of human rights law, in reality characteristics such as nationality or formal legal status can significantly affect the extent of rights an individual is actually accorded... [O]nly the Refugee Convention creates a mechanism – refugee status – by which [rights] attach, and which does not permit derogation”: id. at 12-13; “... [T]he beneficiaries of complementary protection are entitled to the same legal status as Convention refugees, given their analogous circumstances and the Convention’s function as a form of lex specialis for persons protected by the norm of non-refoulement”: id. at 197.
 Id. at 212.
 At one point, McAdam suggests that “... it would in any case be futile for instruments like the CAT to enumerate the legal status arising from the application of non-refoulement since the Refugee Convention (as the lex specialis) already provides an appropriate legal status for any person protected by that principle”: id. At 209-210. Putting to one side the mis-characterization of the Refugee Convention as lex specialis, there is surely no prohibition against other treaties providing for differently defined protected status of their beneficiaries. The Convention relating to the Status of Stateless Persons, for example, does precisely that (a point ironically noted by McAdam, id. at 212). Yet McAdam goes so far as to insist that the fact that Art. 1(A)(1) of the Refugee Convention – which assimilated pre-World War II so-called “statutory refugees” to Convention refugees for purposes of rights entitlement – “... mandates against the creation of additional statuses for persons in need of international protection who do not fall within the Convention definition... (emphasis added)”: id. at 210.
 McAdam rightly invokes the intention of the drafters of the Refugee Convention that the treaty “[e]xpress[ed] the hope that the Convention relating to the Status of Refugees will have value as an example exceeding its contractual scope and that all nations will be guided by it in granting so far as possible to persons in their territory as refugees and who would not be covered by the terms of the Convention, the treatment for which it provides [Final Act of the Conference of Plenipotentiaries, Rec. E]”: id. at 11. But that hortatory statement is no basis to insist that this is a binding obligation. Cf. McAdam, Complementary Protection, at 209, referring to the “... Convention’s function as a ‘charter of minimum rights to be guaranteed to refugees,’ which the drafters envisaged would extend to additional groups of refugees (emphasis added).”
 “The maxim lex specialis derogat lex generali is usually dealt with as a conflict rule. However, it need not be limited to conflict. In the Neumann case, the European Court of Human Rights observed that the provision on compensation in case of unlawful arrest... was not lex specialis in relation to the general rule on compensation... The former did not set aside the latter but was to be ‘taken into account’ when applying the latter. In both cases – that is, either as an application of or an exception to the general rule – the point of the lex specialis rule is to indicate which rule should be applied. In both cases, the special, as it were, steps in to replace the general”: International Law Commission (ILC), “Fragmentation of International Law: Topic (a): The function of the lex specialis rule and the question of ‘self-contained regimes’: An outline” (2005) (“Fragmentation”), at 4.
 G. Fitzmaurice, The Law and Procedure of the International Court of Justice (1986), Vol. I, at 371.
 The exception may be in relation to what are usually referred to as “self-contained regimes” – in Pauwelyn’s view, for example, including WTO law. Where a self-contained regime exists, lex specialis may completely oust the application of more general norms: J. Pauwelyn, Conflict of Norms in Public International Law (2003), at 488. But McAdam clearly does not view refugee law as lex specialis in this strong sense. Rather, she correctly argues that “[h]uman rights law not only provides an additional source of protection for persons with an international protection need, but also strengthens the status accorded to all refugees through its universal application”: McAdam, Complementary Protection, at 253; and “[s]ince universal human rights law is coextensive with Convention status, it follows both as a matter of principle and of law that Convention status should not be used to read down rights. Rather, where human rights law provides more favourable standards, these should be interpolated to improve Convention rights”: id. at 11.
 ILC, “Fragmentation”, at 7.
 V. Conde, Dictionary of International Humanitarian Law (1999), at 82.
 Legality of the Threat or Use of Nuclear Weapons (Advisory Opinion),  ICJ Rep 226, at 240.
 Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory (Advisory Opinion),  ICJ Rep 136, at 178, para. 106.
 By way of an example of an argument invoking lex specialis in order to ground the continued relevance of specialized norms despite the subsequent development of less generous but broader norms, see A. Edwards, “Crossing Legal Borders: The Interface Between Refugee Law, Human Rights Law and Humanitarian Law in the ‘International Protection’ of Refugees,” in R. Arnold and N. Quenivet eds., International Humanitarian Law and International Human Rights Law: Towards a New Merger in International Law 421, at 429 (2008).
 H. Grotius, The Rights of War and Peace (1625), translated by A.C. Campbell, Book II, Ch. XVI, s.XXIX, at 183.
 In any event, the result of such a characterization would not be to give rights to additional classes of non-removable non-citizens, but rather to deny to refugees the benefit of generic human rights entitlements, a position which McAdam clearly does not support: McAdam, Complementary Protection, at 11, 253. “... [T]he [Refugee] Convention is not to be treated as lex specialis enabling one to restrict the implications of general human rights law”: P. Mathew, “Review: J. Hathaway, The Rights of Refugees under International Law,” (2008) 102(1) American Journal of International Law 206, at 207.
 “... [T]he concept of a ‘source’ of a rule of law is important, since it enables rules of law to be identified and distinguished from other rules (in particular from rules de lege ferenda)...”: Jennings and Watt, Oppenheim’s, at 23.
 See Hathaway, Rights of Refugees, at Chapter 1.1.5.
 R. (European Roma Rights Centre and others) v. Immigration Officer at Prague Airport,  EWCA Civ 666 (Eng. CA, May 20, 2003), at para. 28.
 Id. at para. 45, citing the decision on preliminary objections in Cameroon v. Nigeria,  ICJ Rep 2, at para. 39.
 Applicant “A” and Anor v. Minister of Immigration and Multicultural Affairs, (1997) 190 CLR 225 (Aus. HC, Feb. 24, 1997) per Dawson J, adopted in R. (European Roma Rights Centre and others) v. Immigration Officer at Prague Airport,  EWCA Civ 666 (Eng. CA, May 20, 2003), at para. 36.
 See text supra at notes 3-7.
 “Universally recognized treaties can be achieved, but they require political will, compromise, and attention to the sensibilities of all perspectives... If the goal is a world legal order, then the attempts to universalize standards, without the participation and consent of states, impede progress rather than promote it”: Kelly, Twilight, at 539-543.