England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE UPPER TRIBUNAL
(IMMIGRATION AND ASYLUM CHAMBER)
Upper Tribunal Judges Perkins and Craig
Appeal Number: PA/11488/2017

		Royal Courts of Justice Strand, London, WC2A 2LL
		22 June 2021

Covid-19 Protocol:  This judgment was handed down remotely by circulation to the parties' representatives by email, release to BAILII and publication on the Courts and Tribunals Judiciary website.  The date and time for hand-down is deemed to be 10am on 22 June 2021.

Lord Justice Warby:-

Introduction
1. The appellant is a national of Iraq. On New Year's Eve 2007, aged 15, she arrived in the United Kingdom with her mother and sisters to join her father, who was already here. In due course, she was granted discretionary leave to remain. During that period, she posted online statements encouraging jihad. These led to a sentence of imprisonment of 3 ½ years. The respondent (the SSHD) decided to deport the appellant and, when she claimed asylum, refused her application. The SSHD has granted her Restricted Leave to remain, on the basis that returning her to Iraq would expose her to a risk of ill-treatment contrary to Article 3 of the European Convention on Human Rights. But the SSHD determined that the appellant's criminal conduct meant that she did not enjoy the protection of the Geneva Convention on the Status of Refugees (the Refugee Convention). This appeal concerns her claim to protection as a refugee.
2. The First-Tier Tribunal (FTT) allowed an appeal against the SSHD's decision to refuse protection. The Upper Tribunal (UT) allowed an appeal by the SSHD. On this second appeal, brought by permission of my Lady, Elisabeth Laing LJ, the appellant contends that the UT was wrong. The first ground of appeal is excess of jurisdiction: the FTT was the primary decision-maker; the UT was not entitled to interfere with the FTT's decision unless it identified an error of law, which it failed to do; it was not enough that the UT disagreed with the FTT decision, which should have been allowed to stand.
3. For the reasons that follow, I would allow the appeal on that ground and restore the decision of the FTT. I would reject the SSHD's interpretation of the UT's judgment, which is that in substance the UT held that the FTT's decision was irrational. If I were wrong about that, I would allow the appeal on the second ground of appeal: that the UT's decision involved a misunderstanding of the evidence and a misapplication of the relevant test. In my judgment, the FTT's decision was rational and in all other respects lawful, and the UT was wrong in law to interfere with it.
The legal framework
4. As a foreign national sentenced to imprisonment for 12 months or more the appellant is a "foreign criminal" within the meaning of s 32 of the UK Borders Act 2007 (UKBA). Section 32(5) of the UKBA requires the SSHD to make a deportation order in respect of a foreign criminal, but s 33(2)(b) provides that this obligation does not apply where removal would breach the United Kingdom's obligations under the Refugee Convention.
5. Generally, the Refugee Convention prohibits the compulsory return of a refugee to a territory where they risk persecution (refoulement). A refugee, for this purpose, is someone who is outside their country of nationality owing to a well-founded fear of being persecuted there for a reason set out in the Convention, and who is unable or owing to such fear unwilling to avail themselves of the protection of that country: Article 1A. It has been accepted by the SSHD throughout that the appellant satisfies this criterion. The issues have been whether the appellant is excluded from the protection of the Convention by Article 1F(c) or falls within the exception to protection provided for by Article 33(2).
6. Article 33(2) provides that protection from refoulement under the Convention may not be claimed by

"a refugee whom there are reasonable grounds for regarding as a danger to the security of the country in which he is, or who, having been convicted by a final judgment of a particularly serious crime, constitutes a danger to the community of that country."

It will be unnecessary to consider this provision in any detail. The SSHD has never suggested that the appellant is a danger to national security. It was suggested that she is a danger to the UK community, but the FTT held that she is not. The UT dismissed the SSHD's appeal against that decision, and there is no further appeal. We are concerned solely with Article 1F(c).

7. Article 1F(c) provides that an individual is excluded from the protection of the Convention altogether if there are

"… serious reasons for considering that … [s]he has been guilty of acts contrary to the purposes and principles of the United Nations".

As this wording makes clear, Article 1F(c) is not concerned with present danger but is entirely backward-looking. There has been no issue at any stage of these proceedings about the right approach to be adopted as a matter of law, when applying this provision. The governing principles are identified in three decisions: Al-Sirri v Secretary of State for the Home Department [2009] EWCA Civ 222, [2009] Imm AR 624 (Al Sirri (CA)); the same case in the Supreme Court, [2012] UKSC 54, [2013] AC 745 (Al-Sirri (SC)); and Youssef v Secretary of State for the Home Department [2018] EWCA Civ 933 (Youssef).

8. In Al-Sirri (SC) [16] the Supreme Court approved the approach set out by the United Nations High Commission for Refugees in a Background Note of 2003, in these terms:

"The article should be interpreted restrictively and applied with caution. There should be a high threshold 'defined in terms of the gravity of the act in question, the manner in which the act is organised, its international impact and long-term objectives, and the implications for international peace and security'. And there should be serious reasons for considering that the person concerned bore individual responsibility for acts of that character."

9. Lady Hale and Lord Dyson went on to say this, at [37-38]:-

"37. The United Nations Security Council has declared that 'acts, methods, and practices of terrorism are contrary to the purposes and principles of the United Nations' and this is repeated in recital 22 to the Qualification Directive. But it has done so in a context where there is as yet no internationally agreed definition of terrorism, no comprehensive international Convention binding Member States to take action against it, and where the international declarations adopted by the General Assembly are headed 'Measures to eliminate international terrorism'. Above all, however, the principal purposes of the United Nations are to maintain international peace and security, to remove threats to that peace, and to develop friendly relations among nations. It is also noteworthy that the CJEU, despite recital 22 to the Directive, consistently referred to international terrorism, when discussing article 12(2)(c) in B and D."

38. In those circumstances, it is our view that the appropriately cautious and restrictive approach would be to adopt para 17 of the UNHCR Guidelines:

'Article 1F(c) is only triggered in extreme circumstances by activity which attacks the very basis of the international community's coexistence. Such activity must have an international dimension. Crimes capable of affecting international peace, security and peaceful relations between states, as well as serious and sustained violations of human rights would fall under this category.'"

10. The Supreme Court also confirmed that this approach is unaffected by domestic legislation on the topic. Although s 54 of the Immigration, Asylum and Nationality Act 2000 would appear on its face to expand the notion of terrorist activity for this purpose, an international instrument such as the Refugee Convention is autonomous; its scope cannot be defined by domestic legislation; and to read s 54 literally would be inconsistent with Article 12(2)(c) of the EU Qualification Directive, which mirrors Article 1F(c): see Al-Sirri (SC) [36], approving the observations of Sedley LJ in Al-Sirri (CA) [28-29]. When pressed on this point in her oral submissions, Ms Patry accepted on behalf of the SSHD that this was the position.
11. In Youssef, this Court considered and applied the principles identified in Al-Sirri (SC) to the case of an Egyptian national resident in the UK, who was said to have disseminated online statements supportive of terrorism over a 10-year period. It was not alleged that he had incited or encouraged any specific piece of violence, and it was accepted that no link could be shown between his conduct and any specific act of terrorism. But in sermons and other material on the internet which had received hits ranging between 12,000 in a week to 80,000 over an undefined period he had glorified Al Qaeda and its past and present leaders, applauded the organisation's international reach and aspirations, particularly attacks on the US, and implicitly encouraged his audience to emulate the leaders of Al Qaeda: see Youssef [2], [10-15]. The UT accepted the case for the SSHD, that the incitement and encouragement of terrorism "in itself" was contrary to the purposes and principles of the UN, and serious enough to cross the Article 1F(c) threshold: [17-21]. The issue before this Court was whether the UT had "considered sufficiently closely and fully the seriousness and impact of Youssef's conduct, and reached proper conclusions on the point": [81]. The Court concluded that it had not done so, and remitted the case to the UT for reconsideration: [87].
12. Having cited from Al-Sirri (SC) [36-40], Irwin LJ (with whom Rafferty and McCombe LJJ agreed) summarised the key points at [83]:-

"There is a high threshold before Article 1F(c) is triggered. The activity must be capable of affecting international peace and security. However, the Court concluded that 'inducing terror in the civilian population or putting such extreme pressures upon a government will also have the international repercussions referred to …'. That is clearly an issue for specific consideration by the relevant court or tribunal. Finally, the question whether such international repercussions may be established by a person plotting in one country to destabilise another is a question of fact. The test is whether the 'resulting acts have the requisite serious effect'. In short, do the relevant acts have the necessary character and the necessary gravity?"

13. Irwin LJ also took account of the UN Security Council's Resolution 2178, dated 24 September 2014 and thus post-dating Al-Sirri (SC). The Security Council expressed grave concern about (among other things) "the increased use by terrorists and their supporters of communications technology for the purpose of radicalizing to terrorism, recruiting and inciting others to commit terrorist acts, including through the internet …". The Resolution underlined the State's obligation to "prevent terrorists from exploiting technology, communications and resources to incite support for terrorist acts" and "to ensure … that refugee status is not abused by the … facilitators of terrorist acts", in all cases acting "in conformity with … international refugee law". (Security Council Resolutions are admissible material when seeking to apply the provisions of the Charter: see Youssef [43-46] citing Al-Sirri (CA) [29] (Sedley LJ), and Al-Waheed v Ministry of Defence [2017] AC 821 [25], [30] (Lord Sumption)).
14. Irwin LJ went on to say this, at [85]:-

"It may be helpful to consider separately the quality of the acts in question, and their gravity or severity. To adopt an illustration which arose in argument, it is easy to conceive of an immature 18-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, it seems to me, 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 criminal proceedings
15. On 8 April 2015, the appellant pleaded guilty to two offences contrary to the Terrorism Act 2006. Count 1 alleged that she published or caused another to publish statements intending to encourage terrorism, contrary to s 1(2)(i). Count 2 alleged the distribution or circulation of a terrorist publication, intending that the effect of her conduct be a direct or indirect encouragement to the commission, preparation or instigation of acts of terrorism, contrary to sections 2(1)(a) and 2(2)(a). All the offending took place between 1 June 2013 and 14 May 2014. The facts were, in brief, that she had disseminated material online, primarily via Twitter, encouraging young Muslims to join or support jihadi groups fighting overseas.
16. On 11 June 2015, the appellant was sentenced by HHJ Wide QC. He had the benefit of a pre-sentence report (PSR) prepared by probation for the assistance of the sentencing judge. The Judge followed the standard sentencing pattern. He described the appellant's behaviour and assessed its seriousness for sentencing purposes, evaluating her culpability and the harm caused or intended. He considered aggravating and mitigating features, and what reduction should be allowed for the guilty plea. He said her account had gone through twists and turns; she had minimised her behaviour and told lies; only belatedly had she pleaded guilty and admitted that she had intended to encourage terrorism. She had expressed remorse, but only at the sentencing hearing, not to the probation officer. There was other mitigation in the form of her age, her troubled upbringing, and her previous good character. Making a modest reduction in sentence for the late guilty plea, the Judge imposed concurrent sentences of 42 months' imprisonment on each count.
17. Aspects of the sentencing remarks have played a central role in the SSHD's decision-making and arguments in this case. It has been common ground throughout that their factual content is to be taken as accurate, as far as it goes. The key passages for the purposes of this appeal are those that refer to the facts of the offending. I have added numbering, for ease of reference. I have made some obvious corrections to the transcript. There are however some passages where the transcription is clearly wrong but I have left it untouched as I am not sure what was said.

"[1] … This material and its dissemination is an important factor in the encouragement of young men and women to travel abroad and engage in acts of terrorism ... It is a matter of great and justified public concern. You were disseminating such material on a massive scale over a period of just short of a year.

[2] An indication of how busy you were in this activity is that on a site associated with Al Qaeda your Twitter account was noted to be one of sixty-six important Jihadist accounts. The material that you were disseminating encouraged young men to go to fight and you now accept that that was your intention. And furthermore, to encourage women to go to support them and indeed to bring up their children in the belief that it is their duty to take up arms, to wage violent jihad and embrace martyrdom. And furthermore, to encourage mothers to be proud of their sons who die as martyrs.

[3] The material included gruesome images of corpses and prisoners about to be beheaded. You intended to encourage this, as you now accept.

[4] … You said, as is the case, that the messages were copy and pasted from other sources and focused on Koranic verses. And you accepted that some related to jihad. … You said you were not a supporter of violent jihad and you were against groups like Al Qaeda. And you denied intending are working to instruct others to become masters. …

...

[6] [In relation to the PSR] you explained to the probation officer that you have set up social media accounts because you are feeling isolated and you felt it would enable you to connect with other people of a similar age, there was very full social contact, in fact, I am told. You were watching a lot of news programmes highlighting the atrocities committed by Islamic state, terrorists, your hometown in Iraq had been particularly affected by terrorist attacks, your cousin's husband had been killed in one of the attacks and … you started to look at other people's Twitter accounts and cut and paste links and articles into your own account for the edification anybody who wanted to follow you.

…

[11] It is now accepted that it was your intention to encourage terrorism and disseminate terrorist application for that purpose.

[12] It is plain that you were not intending to engage in the reserve yourself and there is no evidence that you are intending to do anything of a purely practical nature. The material, other than some responses to which I have been referred, was not created by you, but forwarded or linked by you. It did not contain practical assistance, though the significance of encouragement should not be minimised, the encouragement to others, both men and women, to fight and to nurture and support those who were fighting.

[13] …

[14] … Coming from a plainly close-knit and concerned family and it has been actually very moving to see your father here, you must be very upset by what you have done and the situation in which you find yourself. I also take into account what seems to me to have been a degree of comparative social isolation and a degree, though it should not be thought that I am putting this very high, a degree of naivete in relation to the consequences of what you were doing. But you were intending what you were intending and that is very serious. …"

The SSHD's decision
18. On 21 January 2016, the SSHD served Notice of Decision to Deport. In reply, the appellant claimed asylum. On 24 October 2017, the SSHD sent the decision letter which is the subject of these proceedings. The decision was that in the light of all the evidence the appellant was "excluded from a grant of asylum as 33(2) and Article 1F applies". It is unnecessary to detail the reasoning in relation to Article 33(2). As for Article 1F(c) issue, the letter quoted the text of the provision. It referred to Al-Sirri (SC) and cited passages from paragraphs [16] and [38] of the judgment in that case. It quoted the words of paragraphs [1-3] of the sentencing remarks of HHJ Wide QC, and concluded that

"… as you have been convicted of a particularly serious crime related to terrorism, to which you were sentenced to 42 months imprisonment (sic), you meet the threshold of Article 1F(c) …. Your conviction and actions … are considered to be contrary to the purposes and principles of the United Nations. You were encouraging people, on a massive scale, to engage in international terrorism. As such there is a clear international dimension to your conduct, which affected international peace and security…"

19. The decision letter annexed the sentencing remarks, the trial record sheet, and some documents relating to an appeal which was evidently refused by the Court of Appeal. It did not refer to or annex the PSR.
The appeal to the FTT
20. The appellant's appeal to the FTT was heard by a two-member Panel (Judges Bird and Norton-Taylor) over three days, at the end of September and in early November 2018. There was oral evidence from a Ms Derosas, and a report and oral evidence from Prof. Silke, all of which went to the appellant's present dangerousness. The 580-page bundle of documents was also largely concerned with the Article 33 issues. The bundle also included the sentencing remarks, the PSR, and three OASys reports assessing the appellant's risks of reoffending at various points in time; but the FTT was not provided with copies of any of the offending tweets or posts, nor any other material (such as a transcript of the prosecution opening of the facts at the sentencing hearing) that went to the issues under Article 1F(c). On 14 January 2019, the FTT promulgated its Decision and Reasons, allowing the appeal in relation to Article 33 and Article 1F(c) for reasons set out over 248 carefully structured paragraphs.
21. Having set out the essential background facts and the decision under appeal, the FTT came at [13-18] to describe "the appellant's offending in more detail". The offences were specified, the statutory provisions were set out, and the sentencing judge's remarks were quoted at length, including but not limited to the passages I have set out above. The FTT made clear that it considered it important to do this for two reasons: "because [the SSHD] places such great emphasis upon them in respect of the case against the Appellant under Article 1F(c), and in part because we have not been provided with the underlying evidence".
22. The Reasons went on to identify the issues, to address matters of law, and to identify the burden and standard of proof, in a fashion that has attracted no criticism. At [36-70], the FTT summarised the content of the evidence before it. At [42] it referred to the PSR, which was evidently a 5-page document that showed the author was "decidedly unimpressed" with the appellant's narrative of events. The author had referred to the appellant being a "prolific poster" on Twitter and Instagram putting up "between 50-60 items a day", a great deal of which were in support of Islamic State. Having then summarised the rival submissions of the parties, the FTT went on at [96-183] to identify its "findings of fact and reasons". Much of this lengthy section of the judgment dealt with facts that go solely or mainly to the Article 33 issues, but it does contain a number of findings that bear on the issues under Article 1F(c).
23. The FTT considered the appellant's offending in the context of her circumstances at the relevant time. It found it likely that the appellant had suffered from mental health difficulties, having witnessed traumatic events in Iraq which included being injured in an airstrike, seeing dead bodies in the street, and witnessing a raid on her grandfather's house by armed soldiers. She had been diagnosed with PTSD, co-morbid depression, and generalised anxiety disorder and appeared to have been socially isolated for most of her life. The FTT said that although these findings did not negate in any way certain other important findings, and in particular the sentencing remarks of Judge Wide QC, they provided a comprehensive picture of the appellant's circumstances.
24. The FTT then turned to the sentencing remarks. In paragraphs [111-112], it recorded the "very significant emphasis" placed upon them by Ms Patry, noting that she had urged the Tribunal "in the strongest terms" not to "go behind" the sentencing remarks. The Tribunal noted that Mr Mackenzie, for the appellant, had acknowledged they should not be disturbed. It stated that, confining itself to the assessment of factual matters, it was not "impermissibly opening a backdoor" to the findings, concluding that "in so far as they go, they represent reliable factual statements." In paragraphs [113-142], the FTT proceeded to conduct a detailed review of the evidence. It made a series of specific findings about the appellant's offending, and her circumstances at the time, identifying the source material on which its findings were based, and how they were arrived at. In doing so, it took the sentencing remarks as its starting point, but sought to read and interpret these in context.
25. It is necessary to describe some aspects of the FTT's findings in a little detail. The key features are these:-

(1) At [115-116], the FTT found that it was likely that some of the posts would have been avowedly pro-Islamic State and some "particularly vile in nature". It did so on the basis of the sentencing remarks and the PSR although – as it observed - the Judge had not made specific findings to this effect, the underlying material was not before it, and it was not clear that this material had been seen by the author of the PSR.

(2) At [118-119], the FTT found that the appellant did not herself commit or intend to commit any acts of terrorism, or intend anything of a practical nature, or provide anything of any practical assistance to others. The vast majority of the material she posted consisted of re-Tweeting content originally posted by others. These findings were based on the nature of the counts, and the content of the sentencing remarks.

(3) At [120] the FTT found that nobody in fact undertook any specific terrorist acts as a result of reading material posted by the appellant. It did so on the basis that there was no evidence to that effect, and the SSHD had not suggested that this was the case.

(4) At [121-125] the FTT addressed the scale of dissemination of the offending material. It found as a fact that over 347 days the appellant posted an average of at most 50 tweets a day, making a total of 17,350 tweets to 8,500 followers. It assessed this as "a large number of posts … over the course of almost a year." The FTT's calculation was done on the basis of agreed figures. The PSR gave figures of 43,000 tweets sent to 85,000 followers, which the FTT described these as "clearly significantly inaccurate". This is plainly right, given that the PSR overstated the scale of publication by a factor of 30. The FTT noted the "real possibility" that the sentencing Judge's description of the volume as "massive" could have been based on the wrong figures, but observed that in any event this was a "highly subjective" description.

(5) At [126-128], the FTT considered the extent to which the material disseminated by the appellant had actually been read, and by whom. It referred to and relied on the description of "How Twitter Works" attached to my judgment in Monroe v Hopkins [2017] 4 WLR 68, and found that it is possible to obtain material indicating how many followers have read particular posts, that such evidence could have been gathered for the proceedings before it; but there was no such evidence. The Tribunal observed that it had no evidence as to who was following the appellant on Twitter.

(6) At [129], relying on the findings in paragraph [2] of the sentencing remarks, the FTT found that the appellant's Twitter account did in fact appear on a website "associated" with Al Qaeda, and that it was in a list of what were described as "sixty-six important jihadist accounts". But it observed that in the absence of any evidence from the SSHD it was unable to make any further findings of fact as to "the nature of the association between the website in question and Al Qaeda; the particular content of the website itself", and other matters relating to the website such as where on the list the appellant's account stood.

(7) At [139-142], the FTT found that the appellant had held extremist views at the time, but that these were not entrenched. It did so on the basis that there was no evidence to indicate extremist views before the offending period, the OASys report of 2015, and its separate finding that the appellant did not currently hold extremist views.

26. At [184], the Tribunal commented on its findings of fact, saying that it had endeavoured to reach clear findings based on the evidence taken as a whole. It emphasised that it had to base its decisions on the material placed before it by the parties, and that "The absence of specific evidence relating to the appellant's social media activities has been something of a recurring theme throughout our findings." The FTT then set out its reasoned conclusions on the Article 33 issues, before coming to the Article 1F(c) issue.
27. At [215], the FTT addressed the relevant legal framework, giving itself the following "core self-directions" in the light of the authorities cited:

"i. given the consequences of the application of Article lF(c), we interpret its provisions restrictively and apply them with caution. This necessarily involves a "high threshold";

ii. the burden of proof showing that the provision applies rests with the Respondent;

iii. in respect of "serious reasons for considering", the threshold is higher than simply reasonable grounds, and there must be something more than mere suspicion. However, the standard is not that of beyond a reasonable doubt;

iv. the phrase "acts contrary to the purposes and principles of the United Nations" has an autonomous meaning. Any definition set out in domestic legislation, whilst relevant, cannot be determinative;

v. the issue of the nature or quality of relevant acts should be assessed separately from their gravity, impact, or severity (for the sake of conciseness, we shall refer hereafter simply to the "quality" and "gravity" of acts);

vi. in assessing the gravity of acts, it is not necessary for the Respondent to show that any particular terrorist act was carried out by any individual as a direct consequence of the Appellant's actions;

vii. ultimately, the assessment of the gravity of acts will be fact-specific."

28. The FTT went on to refer to Al-Sirri (SC), citing the passages from paragraphs [16], [37] and [38] that I have set out above. The FTT's account of the legal principles was described as "impeccable" by Mr Mackenzie in his submissions for the appellant on this appeal. It was not criticised by the UT, nor has it been criticised by Ms Patry. It seems to me to be a neat and accurate encapsulation of all the main points of relevance that emerge from the relevant authorities. Among other things, it aptly identifies the distinction between the nature and quality of acts on the one hand, and their impact or effect on the other.
29. Before the FTT Mr Mackenzie had conceded - as he has on this appeal - that the nature and quality of the appellant's conduct were such as to bring her within the scope of Article 1F(c). Accordingly, the FTT identified the crucial issue as "whether the acts reached a sufficiently high level of gravity for the appellant to be excluded from the protection of the Convention", an issue that fell to be determined on the basis of the evidence adduced before the FTT: [218-219]. At [220], the FTT identified its overall conclusion: "that the Respondent has failed to show that the Appellant's acts were sufficiently grave in terms of their impact upon international peace, security and peaceful relations between states." It went on at [221-239] to identify the matters on which it based that conclusion, making clear that they were all to be read in conjunction with one another. At [240], the FTT reiterated its conclusion that, in a case "that we have decided very much on its own facts", the appellant was not excluded.
30. The FTT's decision is closely reasoned, but for present purposes its key features can be summarised in this way. When deciding such an issue, much will depend on the actual or potential consequences of the acts in question, assessed on a fact-specific basis; so, although the fact that the appellant intended to encourage terrorism and held extremist views was relevant, and counted against the appellant, it was not decisive. There was a large amount of Twitter activity, which should be given some weight, but sheer volume would not of itself satisfy the high threshold. The number of followers was not in itself of great significance; all depends on the facts of the case, as elicited from the evidence. Here, in contrast to Youssef, there was no evidence as to who viewed the posts, no evidence that anybody of influence was commending the appellant, approving her actions, or using her materials, or that her actions led to any terrorist acts, or had any other impact on the behaviour of others, here or abroad.
31. As for the website associated with Al Qaeda, the absence of background evidence had "significantly limited our ability to assess this issue in any more detail". There was no evidence about the links between the website and the organisation, who established the website, how many visitors had been to the site, and so on. This was not a particularly strong factor in the SSHD's favour. There was no evidence that the appellant's posts had in fact led to young people travelling abroad and engaging in acts of terrorism, and the Judge's sentencing remarks could not be taken as a finding to that effect. Unlike the appellant in Youssef, the appellant had not put herself forward as a "scholar" or authoritative voice. Considering the risk that people would read her posts and act on them, a comparison was to be drawn with the comments of Irwin LJ in Youssef [85] about the immature teenager. Although there was more to this case than the actions themselves, most notably the volume of posts and the website reference, these were not sufficiently strong, alone or in combination, to meet the high threshold.
The appeal to the UT

The appellate jurisdiction of the Upper Tribunal

32. The role of the UT is supervisory. An appeal lies, with the permission of the FTT or the UT itself, "on any point of law arising from a decision made by the [FTT]": s 11, Tribunals Courts and Enforcement Act 2008 (TCEA). If the UT finds an error of law it may set aside the decision and remake it: s 12 TCEA. If there is no error of law, the FTT decision will stand. Commonly, the suggestion on appeal is that the FTT has misdirected itself in law. But it is not an error of law to make a finding of fact which the appellate tribunal might not make, or to draw an inference or reach a conclusion with which the UT disagrees. The temptation to dress up or re-package disagreement as a finding that there has been an error of law must be resisted. As Baroness Hale put it in SSHD v AH (Sudan) [2007] UKHL 49 [30]:-

"Appellate courts should not rush to find such misdirections simply because they might have reached a different conclusion on the facts or expressed themselves differently."

33. And as Floyd LJ said in UT (Sri Lanka) v SSHD [2019] EWCA Civ 1095 [19]:

"… although 'error of law' is widely defined, it is not the case that the UT is entitled to remake the decision of the FTT simply because it does not agree with it, or because it thinks it can produce a better one. Thus, the reasons given for considering there to be an error of law really matter."

The grounds of appeal

34. The reasons given in the SSHD's grounds of appeal in relation to the exclusion issue were that the FTT erred in that it

"… [a] went behind the remarks of the sentencing judge, who had already made findings on the gravity of the conduct: he had concluded that AAE had retweeted jihadist material on a "massive scale", that she had been included on a website of important jihadist accounts and that she had intended her acts. All these issues go to the gravity of the acts, but [b] the FTT appears to have focussed almost exclusively on the consequences of the tweeting, rather than the fact that the acts in and of themselves were of sufficient gravity. This amounts to an error of law."

I have added the lettering, to reflect the point, agreed before us, that on analysis this way of putting the matter advances two separate grounds of appeal.

35. In the light of Floyd LJ's observations in UT (Sri Lanka), some short comments on the Grounds of Appeal are appropriate. The first is that they contain no express complaint that the FTT misdirected itself in law. Secondly, the notion of "going behind" the sentencing remarks is a metaphor, which is not a term of art, and seems to me potentially problematic. I have wondered about its status as an error of law. Criminal sentencing is undertaken in separate proceedings, on a legally different basis from the assessment of gravity for the purposes of Article 1F(c). Here, it was sensibly agreed that the sentencing remarks could be relied on for their factual findings. But that may not always be agreed, in which case issues might arise as to the legal basis for asserting that the remarks bind the appellant. In any event, factual assessments may not always be crystal clear, or their precise basis may be debatable. Sentencing remarks may quite properly contain language such as (in this case) "massive scale", which is evaluative rather than factual in nature. The boundary between fact and comment may not always be obvious.
36. Thirdly, it seems to me that the Grounds could have more clearly identified why it was an error of law to focus "almost exclusively" on the consequences of the appellant's acts. The Grounds did not assert that the consequences were irrelevant, nor could they have done so in the light of Youssef. But nor did the Grounds assert that the acts themselves were a relevant factor which the FTT ignored.
37. Permission to appeal was refused by the FTT (Judge Grant-Hutchison) but granted by the UT (UTJ Chalkley). Permission to appeal on the Article 1F(c) issue was granted on the basis that "… it is arguable that the panel …may have erred in law by going behind the sentencing remarks … when considering the level of gravity of the appellant's actions" (emphasis in original). This reflects what I have labelled ground [a]. The UTJ did not expressly identify ground [b] as a possible error of law, though it is fair to say that it was not separately identified at the time, and the UTJ expressly stated that the SSHD had permission to appeal "on all grounds."

The UT decision

38. In the introductory section of their reasoned decision, at [7], the UT reminded themselves that "we are not, at least initially, primary decision makers. Our function is to see if the [SSHD] is able to show that the [FTT's] decision to allow the appeal was wrong in law…"
39. The UT addressed the SSHD's appeal on the Article 1F(c) issue between paragraphs [42] and [71]. They began by summarising the grounds of appeal, and the arguments advanced by the SSHD. They recorded submissions by Ms Patry as follows: (1) the FTT's finding that the acts committed did not reach a sufficient level of gravity "was based on an unlawful rejection of the finding of the sentencing judge" that the appellant had "re-tweeted jihadist material on a 'massive scale', that she had been included on a website of important jihadist accounts and that she had intended her acts"; (2) the FTT was wrong to "give weight to the fact that there was no evidence of re-tweeting"; (3) the FTT should not have "stepped away from" the sentencing judge's "observation" that the dissemination of material was "an important factor in the encouragement of young men and women" to travel abroad and engage in terrorism; (4) "more regard should have been given" to the findings of the sentencing judge about the gravity and severity of the acts, and his use of the phrase "massive scale"; and (5) it is not necessary to look for evidence of anyone taking any notice.
40. Having summarised the submissions for the appellant, the UT encapsulated how they saw the issue at [55]:

"… it comes down to this. It is the Secretary of State's contention that the Tribunal was not entitled to find on the evidence before it that the high threshold had not been reached. The parties agree that the Claimant has committed terrorist offences and that she has been involved in tweeting profoundly unpleasant material. There is some disagreement about the extent of its circulation but on anybody's version it has gone to thousands of people. There is no evidence that anybody has taken any direct notice of it and certainly no evidence that anybody has been actually inspired to do anything of a serious nature as a result of reading it."

41. The UT reminded themselves of the legal principles, noting that "all of these things were fully in the mind of the First-tier Tribunal", and then went on to say this:

"62. The nature of the conduct is not only clearly contrary to the purposes of the United Nations but is an example of a relatively recent way in which those purposes can be thwarted.

63. There are three factors in the First-tier Tribunal's reasoning which, we find, do at least cumulatively, make out the necessary gravity.

64. First is the volume of Twitter activity. The First-tier Tribunal asserted that that on its own is not of "great significance". According to the First-tier Tribunal the problem is not the scale of the dissemination but the effect that it has. We agree in the sense that we doubt that volume alone could establish the necessary gravity. Drivel, on a large scale, makes a lot of drivel but would not undermine the purpose of the United Nations. The conduct complained of here was not "drivel". It included "gruesome images of prisoners about to be beheaded" which conduct the Claimant intended to encourage (see sentencing remarks). We regard the scale as important. The conduct complained of was not a one-off rant said in private and possibly overheard but was circulated deliberately to as large an audience as could be found for the purpose of encouraging terror. As indicated there is some dispute over the size of that audience but it was conduct intended to encourage terrorism and it was sent to thousands of people.

65. Second, there is the timescale. According to the sentencing remarks this went on for nearly a year. Again, contrary to the view of the First-tier Tribunal, we find that this is a significant aggravating factor that increases the gravity of the conduct. It was not someone "sounding off" but persistent as well as large scale misconduct.

66. We note the First-tier Tribunal was not able to get a clear understanding of the number of occasions offensive tweets were sent. It rejected the highest figures found in some of the papers and gave good reasons for that but it is also clear that the First-tier Tribunal found expressly that the claimant "posted what we consider to be an average of at most 50 tweets a day over the course of 347 days" (see paragraph 121). We find that these are levels of activity which take the case away from the isolated youth ranting and elevate it into a higher category and that the First-tier Tribunal was just wrong to take a contrary view.

67. Further, these two points are linked by supporting a criminal prosecution. We are not saying that a criminal conviction is necessary or determinative and we accept that concepts and even phrases that appear in national criminal legislation and international treaties should not be assumed always to mean the same in each context. Nevertheless the fact that conduct amounted to offences under the Terrorism Act sufficiently serious to warrant a substantial sentence of imprisonment is, we find, itself a strong indicator that the conduct was contrary to the purposes of the United Nations and the First-tier Tribunal, we find, rather lost sight of this.

68. We also do find it significant that the claimant's website was "on a site associated with Al Qaeda your Twitter account was noted to be one of 66 important jihadist accounts". We would have liked to know more about that as no doubt would the First-tier Tribunal but, unlike the First-tier Tribunal, we find that this clear, if limited, description is enough to show this claimant's activities were of sufficient prominence to be commended by Al Qaeda and the First-tier Tribunal was, with great respect, wrong to say that this was not a factor singularly or cumulatively to elevate the conduct to be sufficiently grave and severe to disqualify the claimant. It is clear evidence that the Claimant's intended behaviour mattered to Al Qaeda. It was significant even though there is no evidence that it was heeded directly."

42. The UT recorded at [69-70] that they had reflected before making these observations, and had applied their minds carefully to the question of "whether we are merely disagreeing with the [FTT], which is not relevant, or determining that they were wrong to take the view that it did which is highly pertinent." They had decided it was the latter. The UT continued as follows:

"71. We find that a person who sets out to encourage terrorism and does so by circulating encouraging and destructive material over the internet many times a day for the best part of a year cannot avoid being found to have been undermining the purposes of the United Nations. The necessarily high threshold is crossed by the repetition of the offences. In this case it is compounded by the clear evidence that there was success in the project because the conduct was commended by an Al Qaeda supporting organisation. That is independent evidence of the importance of the conduct. The First-tier Tribunal should have given a lot more weight to these things. This is why we find the First-tier Tribunal was wrong and why we overturn it.

…

78 … we have explained why we disagree with the First-tier Tribunal."

Assessment
43. The heart of the UT's reasoning is to be found in paragraphs [62-68] and in its own summaries at [71] and [78]. A number of points can be made about these passages of the decision.

(1) The language is that of fact-finding, coupled with assertions that the FTT was wrong. Repeatedly, the UT states that "we find" a certain proposition to be made out. Nowhere it is said in terms that the FTT made any error of law.

(2) The Tribunal does not state that it upholds any part of the SSHD's grounds of appeal, which are not referred to here. The UT does not speak of the FTT "going behind" the sentencing remarks, or say anything to that effect. It does not appear to have upheld ground of appeal (a). The contrast drawn here is not between the sentencing remarks and the FTT decision. It is between the FTT decision and the facts as found by the UT. Nor does the UT, in terms, adopt the SSHD's ground (b).

(3) The ways in which the FTT is said to have gone "wrong" include taking a different view from that of the UT on (a) the importance of the scale or volume of the appellant's Twitter activity; (b) the significance of the duration and persistence of her conduct; (c) the fact that the facts "supported a criminal prosecution"; and (d) the cumulative significance of these factors.

(4) On the face of the UT's decision in this case, the difference between concluding that the FTT was wrong and disagreeing with it looks like a distinction without a difference.

(5) Summarising its criticisms, the UT itself identified the FTT's error as failing to "give a lot more weight to these things". Summarising its decision, the UT said "we disagree" with the FTT.

44. Language such as this bears the hallmark of appellate error. It is trite, but relevant, to say that mere disagreement is not a ground on which to reverse a decision. The assessment of weight is generally for the court or tribunal of first instance. All of this lends strong support to Mr Mackenzie's submission that the UT departed from its own self-direction; it substituted its own decision for that of the primary decision-maker, having failed to identify any error of law in the decision of the FTT. Reviewing a multi-factorial assessment, which has been categorised as a question of fact, it reached its own conclusions on individual elements and on the significance of those elements in the context of the applicable law, and then substituted those conclusions for those of the FTT.
45. Ms Patry points out that we are not construing a statute, and urges us not to take too strict or semantic an approach to the interpretation of the grounds of appeal or the UT decision. She submits that the substance and reality of the position is that when it said the FTT was "wrong" the UT meant that it was "wrong in law." The SSHD's ground (b) was a contention that the FTT decision was irrational, and it was that contention with which the UT agreed, and rightly so.
46. This argument faces a number of obvious difficulties. One of them is that the word "irrational" is not to be found anywhere in the relevant ground of appeal, the skeleton argument in support of that ground of appeal, the UT's account of Ms Patry's submissions, or the UT's own reasoning on the Article 1F(c) issue. I do not believe that in saying this I am taking too narrow or literal an approach. The term is very familiar, as a threshold for appellate intervention: see, for instance McCloskey J in SSHD v Greenwood [2015] UKUT 00629 (IAC) [17]: "The touchstone for intervention is irrationality. This Tribunal can find an error of law in the context of this appeal only if the outcome of the application of the correct legal test is vitiated by irrationality." I also note that the word irrational was used by the SSHD in grounds of appeal 3 and 4, and in Ms Patry's skeleton argument on those issues. Other formulae that bear the same meaning are also familiar. Some of them were mentioned by McCloskey J in the passage cited. One is, did the conclusion "fall within the band, or range, of conclusions reasonably open and available" to the first-instance decision-maker. That formula is not to be found in this part of the UT decision.
47. Of course, this is not just a question of language. The issue is whether, in substance, the UT did or did not find an error of law. Ms Patry points to the UT's self-direction at [7] and its self-enquiry at [69-70] as indications that the Tribunal had the limits of its jurisdiction and its function clearly in mind. She highlights paragraph [55], where the issue for decision is identified as whether the FTT was "entitled" to reach the conclusions that it did. And I note that in paragraph [71] the UT found that a person who had acted as it found the appellant had acted "cannot avoid being found to have been undermining the purposes of the United Nations" (my emphasis). These features of the decision do provide some support for Ms Patry's skilful argument.
48. Having reflected on these submissions, however, I cannot accept Ms Patry's analysis. There are three main reasons for that. The first is that this is not how the SSHD's case was expressed. As I read the grounds of appeal, and the way it was put in the written and oral argument, the SSHD's case was twofold. It was argued that even if the appellant's conduct had no significant effect or impact her behaviour, properly assessed, was bad enough to reach the threshold of gravity "in and of itself"; the FTT had failed to give enough weight to the appellant's conduct as opposed to its effect. And it was said that in assessing gravity, both as to conduct and effect, the FTT was bound to follow the findings of the sentencing judge and failed to do so. Much was said about the FTT paying insufficient regard, or giving too much or too little weight, to certain factors. These are not irrationality arguments. So I do not consider that the issue of whether the FTT's decision was irrational was an issue raised for decision by the UT.
49. Secondly, although we should doubtless read the decision of an expert appellate tribunal with a degree of benevolence, we must attach some significance to the language it uses to explain its reasoning. Not only does the UT make no reference to the test of irrationality, or cognate tests, it does use the language of "weight" and "disagreement" that I have quoted above. So whatever it may have meant by the wording in [55] and [71], it cannot be said that the UT's analysis was conducted within the framework of the irrationality principle. The argument for the SSHD entirely fails to account for these striking features of the decision. At best, as I see it, the UT's decision was an uneasy mix of language redolent of appellate over-reach with a sprinkling of wording that is consistent with a proper recognition of the UT's jurisdictional boundaries.
50. Thirdly, I do not believe that the substance of the UT's reasoning can be said to reflect the proper application of an irrationality test. At [63], the UT found that there were three factors that "do, at least cumulatively, make out the necessary gravity" (my emphasis). It is not immediately obvious what the third factor was, but I would accept Ms Patry's analysis: the three factors were the volume or scale of the tweeting ([64]), the timescale/levels of tweeting ([65-66]), and the fact of a criminal prosecution and substantial sentence ([67]). The words I have emphasised above reflect an acceptance that there was at least room for debate or disagreement about whether any one of those factors would be sufficient in itself. The same assessment is reflected in the UT's analysis of the three factors individually. To say, for instance, that the scale of the tweeting is something that "we regard … as important" is a far cry from saying that the tweeting was on such a scale that it reached the threshold by itself, and that no other view could rationally be taken. Similarly, the UT's finding that the sentence of imprisonment imposed on the appellant was "itself a strong indicator" that the threshold was reached cannot be read as finding that this would have been enough. So, the real nub of the UT's decision-making in these paragraphs is that the three factors taken together "do" make out the necessary gravity. That, clearly, is not at all the same as finding that they "must" do so, and that no other conclusion is available to a reasonable tribunal.
51. Paragraph [68] of the UT's decision shows that they also took account of a fourth factor: the website "associated with" Al Qaeda. The Tribunal held that the description given in the sentencing remarks was enough to make this a matter sufficient "singularly or cumulatively to elevate the conduct to be sufficiently grave and severe to disqualify the claimant", and that the FTT was "wrong" to say otherwise. Given that the UT had already pronounced that three factors taken together were enough to cross the threshold, this additional passage complicates analysis of its reasoning. But on careful scrutiny, I believe three points are clear. The UT regarded the website as an additional factor, not one that was essential or integral to the reasoning set out at [63]. This is clear from the statement that this factor "compounds" the position. Secondly, and contrary to the submission of Ms Patry, paragraph [68] cannot be read as a finding that the website factor of itself obliged the FTT to find that the appellant was disqualified from refugee status. That follows from the UT's use of the words "singularly or cumulatively". And finally, whilst the UT said the FTT was "wrong" on this point, it did not say that the only available conclusion was that this and the other three factors taken together crossed the threshold.
52. For all these reasons, I conclude that the appellant's first ground of appeal is made out. The UT did not make any finding that the FTT erred in law and accordingly it had no jurisdiction to interfere with the FTT decision.
53. If it were necessary, I would have upheld the appellant's second and alternative ground of appeal. I am quite satisfied that if, contrary to my analysis, the UT's decision was that the FTT reached an irrational conclusion, the UT was wrong in law. Ms Patry submitted to us that if we got to this point we should show some deference to the UT's decision, allowing it a margin of appreciation. That is not how we approach an issue of this kind. The reasons are obvious. Indeed, reflecting on Ms Patry's submission, it seems to me to reveal the paradox: if there is more than one possible view about the rationality of the FTT's decision, then that decision cannot be held to be irrational. But there is more to it than that.
54. The UT was wrong in law to proceed, if it did, on the basis that the FTT was obliged to make a finding of the necessary gravity on the basis of the appellant's acts in and of themselves, without regard to the impact of those acts. The strand of reasoning that begins at [63] ignores the impact of the judgment in Youssef, and its application to the facts of this case. The UT was also clearly wrong in law in its approach to the website issue, to which it evidently attached considerable weight. The UT referred at [68] to "the [appellant's] website", but her activity was on Twitter and Instagram. And there was no evidence that – contrary to the express finding of the FTT - the appellant's Twitter activity had been "commended by Al Qaeda". The SSHD had never submitted that this was so. The sentencing remarks cannot sustain any such finding. All they revealed was a link between the appellant's Twitter account and a website which was "associated with" Al Qaeda in some way that was not disclosed. As Mr Mackenzie submitted, this was an instance of the UT going behind or – as I would prefer to put it – beyond the sentencing remarks. The FTT decision on this point as on the others was a better reflection of the sentencing remarks and what could and could not be drawn from them, and it was not irrational.
55. In summary, the UT's decision was in substance an impermissible interference with the rational evaluation of an expert tribunal which correctly directed itself as to the law, took all relevant factors into consideration, and did not leave any relevant factor out of account. The appeal should be allowed.

Lord Justice Birss:-

56. I agree. For the reasons explained by my Lord, this appeal should be allowed. The UT is only entitled to overturn a decision of the FTT if an error of law by the FTT is identified. The only error of law which counsel for the SSHD contended for as a basis for the UT's decision was that the FTT's decision on Article 1F(c) was irrational. However as my Lord has shown, the UT did not put its decision on that basis and, since no other error of law was identified, the UT had no jurisdiction to overturn the FTT. Furthermore, the FTT's decision was one which was open to it on the evidence, such as it was. Therefore the FTT's decision was not an irrational decision. Accordingly even if the UT really had decided to overturn the FTT on the ground that the FTT's decision was irrational, the UT would have been wrong in law to do so. Thus the appeal succeeds.

Lady Justice Elisabeth Laing:-

57. I agree with both judgments.