Bad, but not THAT bad: the Court of Appeal in Youssef and N2

The key question for the Court of Appeal in these linked cases was whether a person can be excluded from protection under the Refugee Convention for having encouraged or prepared for acts of terrorism – without having committed or attempted to commit a specific act, or being linked to any specific act by others. In other words – how bad is that bad?

The cases

The facts of the cases were both straightforward and striking. Both appellants had sought refugee protection in the UK. The Secretary of State had decided it was unavailable to either, as the appellants fells within the “exclusion clause” in Article 1F of the Refugee Convention – and specifically within Article 1F(c), which excludes any individual in respect of whom there are “serious reasons for considering” that he or she “has been guilty of acts contrary to the purposes and principles of the United Nations.”

What were the acts relied on by the Secretary of State?

Exclusion from refugee protection is, of course, an exceptionally serious matter

• The first appellant, Mr Youssef, had published a large number of sermons and other writings on his website glorifying Al-Qaeda and its leaders, including Osama bin Laden, and encouraging readers to follow their example. The Upper Tribunal had found as a matter of fact that, by his writings, Youssef had knowingly incited and encouraged – albeit in a broad and general way – acts of terrorism.
• The second appellant, N2, had entered the UK on a false passport; been refused asylum; and subsequently received a nine-year prison sentence for offences contrary to the Terrorism Act 2000. The offences related to the possession of computer files containing directions on how to establish a jihadist organisation and “genuine instructions” on the making of bombs and detonators. N2 had been sentenced on the basis that he appeared to be a “sleeper” for an unknown terrorist organisation, though he did not appear to have been involved in planning a specific attack.

In both cases, the Upper Tribunal had agreed that the acts in question fell within the ambit of Article 1F(c).

The framework

Exclusion from refugee protection is, of course, an exceptionally serious matter. Although the Court of Appeal was at pains to point out that its decision would not mean either appellant faced “immediate removal or deportation from the UK” (as both had been granted six months’ restricted leave to remain), the baseline position is that falling within Article 1F means a person who might otherwise be recognised as a refugee – and who may therefore face a real risk of persecution on return to his or her country of nationality – can be returned to that country despite the risk.

For precisely this reason, Article 1F is targeted at relatively confined group: those “with respect to whom there are serious reasons for considering that” –

• they have “committed a crime against peace, a war crime, or a crime against humanity, as defined in the international instruments drawn up to make provision in respect of such crimes”;
• they have “committed a serious non-political crime outside the country of refuge prior to [their] admission to that country as a refugee”; or
• they have “been guilty of acts contrary to the purposes and principles of the United Nations.”

Although there have (perhaps unsurprisingly) been some domestic attempts to stretch the scope of subsection (b), it is clear on the face of the article that each limb is directed at a particularly grave form of misconduct. Acts which are merely “bad” – even criminal – will not suffice.

What, then, of the appellants? Were the acts they had committed that bad?

The arguments

Three main arguments were made as to why the appellants fell outside the scope of Article 1F(c).

1. The first was that to fall within the subsection, an act must either constitute a crime under Article 1F(a), or must be shown to have led to the commission of such a crime. Neither was the case here.
2. The second was that, at least, Article 1F(c) required a completed or attempted terrorist act – not mere preparatory acts.
3. The third was that, in any event (and in the case of Youssef only), “mere speech” could not be contrary to the principles and purposes of the United Nations, and in this case was not sufficiently serious to engage Article 1F(c) – an issue the Upper Tribunal had entirely failed to consider.

The decision

The Court of Appeal (in a leading judgment by Irwin LJ) rejected the first argument for fairly comprehensible reasons: the proposed reading of Article 1F(c) would render it completely superfluous, and would ignore the difference between its reference to “acts” and the references in (a) and (b) to “crimes”.

The reasoning in relation to the second argument was more limited: the Court simply identified a UN Security Council resolution and a decision of the CJEU which both, in its view, supported the conclusion that Article 1F(c) was not confined to completed or attempted terrorist acts. Rather, it seems the Court considered (at least implicitly) that the degree of connection with specific acts would go to the fact-sensitive question of whether the identified conduct was “contrary to the purposes and principles of the United Nations” and was sufficiently serious to engage the exclusion clause (see further below).

In respect of the third argument, the Court took its lead from the judgment of the Supreme Court in Al-Sirri v SSHD [2012] UKSC 54. In that case, the Supreme Court had:

• endorsed para 17 of the UNHCR guidelines on Article 1F(c), which state that the clause is “only triggered in extreme circumstances by activity which attacks the very basis of the international community’s coexistence”, and only by activity which has an “international dimension”;
• considered it “very likely” that “inducing terror in the civilian population or putting such extreme pressures upon a government [as terrorist acts are intended to] will… have the international repercussions referred to by the UNHCR”; and
• considered that, in cases where “a person plots in one country to destabilise conditions in another”, whether Article 1F(c) is engaged will depend on the circumstances – the overriding question being “whether the resulting acts have the requisite serious effect upon international peace, security and peaceful relations between states.”

On this basis, the Court of Appeal summarised the test under Article 1F(c) as requiring both “the necessary character” and “the necessary gravity”. It concluded that, in Youssef’s case, the Upper Tribunal had failed to give proper consideration to the “gravity” element; it therefore allowed the appeal on this narrow ground, and remitted the matter for reconsideration.

Points of interest

Perhaps the most interesting aspects of the judgment are the issues which were not, in fact, decided. I have identified three.

How do we tell what is really that bad?

The first is the question ultimately remitted to the Upper Tribunal: namely, whether the acts committed by Youssef were serious enough to fall within Article 1F(c) even where they could not be linked to the commission or attempted commission of specific terrorist acts.

This is particularly interesting because, in considering how a court or tribunal might assess the “gravity” limb of the test, the Court have the following example:

… it is easy to conceive of an immature 19 year old going online from his suburban bedroom, and using the most lurid terms in calling for international jihad. The nature or quality of this would… satisfy the requirements of Article 1F(c). It would represent active encouragement or incitement of international terror. However it would be unlikely, without more, to be grave enough in its impact to satisfy the approach laid down in Al-Sirri. That might well require more: evidence of wide international readership, of large-scale repetition or re-tweeting, or citation by those who were moved to join an armed struggle, for example.

The Court considered it “obviously right” that “careful consideration is given to the gravity or impact of any acts relied on” (emphasis added).

The question for future cases seems likely to be how widely the net will be cast when looking for evidence of impact.

In so reasoning, the Court of Appeal appears to be treating impact as the key metric for assessing gravity – indeed, the factors identified in its example are all effectively proxies (albeit imprecise ones) for the impact of the hypothetical statements. This approach seems broadly appropriate in light of the Supreme Court’s focus on the effect of the relevant acts (see [16.3] above). However, it is difficult to see how the kinds of indicator identified by the Court of Appeal – particularly readership and repetition – could alone be capable of demonstrating a “serious effect upon international peace, security and peaceful relations” (as required by the Supreme Court). Similarly, it is not entirely clear how the focus on impact fits with the Court’s dismissal of N2’s appeal – given that no evidence of impact seems to be identified anywhere in the judgment. The question for future cases seems likely to be how widely the net will be cast when looking for evidence of impact.

What does the UN Security Council tell us about the 1F(c) threshold?

The second point of interest arises from the Court of Appeal’s use of resolutions of the UN Security Council. These the Court drew on not only to determine whether general incitement or preparation could be “contrary to the principles and purposes of the United Nations”, but also – and perhaps more controversially – to determine the threshold of gravity required to engage Article 1F(c). Thus, the Court of Appeal found that the inclusion in resolutions of language condemning incitement to terrorism, and requesting States to avoid providing “safe haven” to those involved in terrorist acts, “clearly point(ed) to consideration of Article 1F, and to obligations on States to take action rather more broadly, and more readily, than would be indicated by a straightforward application of paragraphs 17 and 18 of [the UNHCR guidance]” (see [16.1] above).

the Court does not appear to have considered the language of the resolutions in their full context

It is clear, of course, that Security Council resolutions are relevant to the interpretation of the Refugee Convention insofar as they constitute evidence of subsequent State practice under Article 31 of the Vienna Convention on the Law of Treaties. Although the Court of Appeal did not expressly recognise this as the relevant interpretive mechanism, it was not wrong to take the resolutions into account. However, there are several reasons to doubt the conclusion it ultimately drew from them.

First, the Court does not appear to have considered the language of the resolutions in their full context. For example:

• Resolution 1624/2005, in its recitals, does condemn “the incitement of terrorist acts” and repudiate “the justification or glorification of terrorist acts that may incite further terrorist acts.” However, the operative paragraph calls upon States to “adopt such measures as may be necessary and appropriate and in accordance with their obligations under international law” to “prohibit by law incitement to commit a terrorist act or acts” and “deny safe haven to any person with respect to whom there is credible and relevant information giving serious reasons for considering that they have been guilty of such conduct” (emphasis added). Thus, the Resolution limits is call to action under Article 1F to individuals guilty of the types of incitement prohibited by domestic law, and makes it clear that such action should be undertaken in accordance with existing international obligations (including, necessarily, under the Refugee Convention).
• Similarly, the only citation of Resolution 2178/2014 is of its preamble, which expresses “concern” over the use of communications technology for the purposes of “radicalizing to terrorism” but also calls on States to:
  • “act cooperatively to prevent terrorists from exploiting technology… to incite support for terrorist acts, while respecting human rights and fundamental freedoms in compliance with other obligations in international law”;
  • ensure “in conformity with… international refugee law” that refugee status “is not abused by the perpetrators, organizers or facilitators of terrorist acts”.
• Thus, the Resolution again emphasises adherence to (not modification of) existing international obligations, and gives examples of abuse (only) by individuals involved in specific terrorist acts.

In consequence, while the resolutions are clearly relevant in determining the character of relevant “acts” under Article 1F(c), there is nothing in the passages cited which can properly be seen as lowering the threshold of gravity which – as recognised by the Supreme Court in Al-Sirri – is inherent in the article itself, and hence which the resolutions urge States in terms to respect.

What can Parliament say about the “one true meaning” of the Convention?

The final point of interest concerns the Court’s comments on the capacity of EU or domestic legislation to dictate the meaning of the Refugee Convention as applied by the courts.

The issue arose because both the Recital to the Qualification Directive and the Immigration Act 2006 contain provisions which purport to bear directly on the interpretation of Article 1F(c).

• Article 22 of the Recital provides that “Acts contrary to the purposes and principles of the United Nations… are, amongst others, embodied in the United Nations Resolutions relating to measures combating terrorism, which declare that… ‘knowingly financing planning and inciting terrorist acts are also contrary to the purposes and principles of the United Nations.’”
• Section 54 of the 2006 Act provides that “in the construction and application of Article 1F(c) of the Refugee Convention the reference to acts contrary to the purposes and principles of the United Nations shall be taken as including, in particular… (b) acts of encouraging or inducing others to commit, prepare or instigate terrorism (whether or not the acts amount to an actual or inchoate offence).”

The Court of Appeal considered, at least briefly, the impact of these provisions on its deliberations. Irwin LJ – noting the appellants’ reliance on the authorities endorsing the need for a “single autonomous meaning” of the Refugee Convention – said he “[did] not see how a single autonomous meaning [could] be preserved if the meaning of the Convention is altered by the European Directive, or by the 2006 Act.” However, it was unnecessary to take the issue further as he had independently reached a conclusion consistent with both instruments.

McCombe LJ was still more taken with this point, and devoted a separate judgment to it. He wished expressly to “hold over for consideration in some future case the precise extent of the power of Parliament and the institutions of the European Union to legislate for the interpretation (in English law) of an international instrument such as the Refugee Convention.” His Lordship suggested – by reference to the decision of the House of Lords in R v SSHD; ex parte Adan [2001] 2 AC 277 – that it must be open to Parliament to do so.

There are difficulties with McCombe LJ’s reading of the relevant passage of Adan; however, at least on a basic view of first principles, his conclusion appears sound. The reasoning would be as follows:

• The UK has obligations at international law under instruments such as the Refugee Convention. However, those obligations have no effect as a matter of domestic law unless, until and to the extent that they are incorporated into domestic legislation.
• Thus, Parliament is free to legislate in ways which are fully, partially, or even completely inconsistent with its international obligations – and the courts (subject only to higher-level constraints, such as directly applicable EU legislation or fundamental constitutional principles) are obliged to give effect to the will of Parliament. This must include giving effect to any definition legislated for by Parliament of a term used in an international instrument.

It is for these reasons that in Australia (to take another common-law example), Parliament has long been putting its own domestic “spin” on critical provisions of the refugee definition – to an extent many would consider incompatible with a proper interpretation of the Refugee Convention[1] – apparently without the batting of a judicial eyelid.[2]

It would appear that the Court of Appeal did not consider the issue so straightforward – an indication which will doubtless be of interest to rights-focused advocates. In addition, the more protective human rights framework in the UK seems likely to place much more meaningful constraints on legislative action than has been the case in Australia. The combination, it is to be hoped, will prove effective in preserving the “one true meaning” of the Refugee Convention in UK domestic law.

Conclusion

The judgment of the Court of Appeal identifies a deceptively simple test

Exclusion from the protection of the Refugee Convention is a serious matter. The judgment of the Court of Appeal identifies a deceptively simple test for the application of Article 1F(c), and provides some initial indications of how it might be applied in practice. However, the judgment arguably goes too far in suggesting that the “gravity” threshold which forms part of the test should be lowered in cases involving incitement to or preparation for terrorist acts. The judgment also raises important questions concerning the relationship between courts and the legislature in the interpretation and application of the Refugee Convention – questions which, in the current political climate, may come before the courts again sooner rather than later.

NOTES:

[1] For example, in 2014 Parliament removed references to the Refugee Convention from the Migration Act 1958 altogether, introducing what was prima facie a purely statutory definition of a “refugee”. In his Second Reading Speech, the Minister for Immigration and Border Protection stated expressly that the purpose of the amendments was to ensure that Australia was not “subject to the interpretations of foreign courts of judicial bodies which seek to expand the scope of the Refugee Convention well beyond what was ever intended by this country or this Parliament.” The Explanatory Memorandum suggested an intention to “codify Australia’s interpretation” of its obligations under the Convention. The new section 5J(4) requires that a Convention reason be “the essential and significant reason” for persecution, and that “persecution” must involve both “serious harm” and “systematic and discriminatory conduct”. In addition, new section 5J(1)(c) requires that the real risk of persecution “relates to all areas of a receiving country” – a fundamental departure from the generally accepted understanding of the internal relocation principle.

[2] Commentators at the time noted that the amendments posed an “unprecedented interpretative challenge” for Australian courts, given that it is “well accepted that primacy is to be given to the text of legislative provisions” and that any ability to use international law to interpret the Migration Act has been deliberately limited: see Burson and Cantor, Human Rights and the Refugee Definition: Comparative Legal Practice and Theory (Brill Nijhoff, 2016) p 84. Since then, at least one Federal Court decision has held that “the absence of any incorporation by reference of [the Refugee Convention] means that, materially, s 5J falls for interpretation not as an international instrument but as a provision in a domestic statute” (at [18]) and that, in consequence, whether the new provisions “codified” or “radically” altered” the internal relocation principle as set out in Januzi (which had previously been followed by Australian courts) was “at least moot” (at [25]): CDZ16 v Minister for Immigration and Border Protection [2017] FCA 967.