Tribunal decisions

The decision

IH (s.72; ‘Particularly Serious Crime’) Eritrea [2009] UKAIT 00012

ASYLUM AND IMMIGRATION TRIBUNAL

THE IMMIGRATION ACTS

Heard at Field House Dates of Hearing: 12 February,
3 and 9 September 2008
Before

Mr C M G Ockelton, Deputy President of the Asylum and Immigration Tribunal
Senior Immigration Judge Lane
Senior Immigration Judge Grubb

Between

IH
Appellant
and

THE SECRETARY OF STATE FOR THE HOME DEPARTMENT

Respondent

Representation:

For the Appellant: Mr P Draycott instructed by Paragon Law
For the Respondent: Mr P Patel instructed by the Treasury Solicitors

The presumptions in s.72 of the Nationality, Immigration and Asylum Act 2002 that in the circumstances specified a person has been convicted by a final judgment of a “particularly serious crime” for the purposes of Art 33(2) of the Refugee Convention if read as irrebuttable are inconsistent with Art 21.2 of the EU Qualification Directive (Council Directive 2004/83/EC) which gives effect to the autonomous international meaning of Art 33(2) as part of EU law. As a consequence, the presumptions in s.72 must be read as being rebuttable.

DETERMINATION AND REASONS

Introduction

1. The Appellant is a citizen of Eritrea who was born on 1 April 1973. He is of mixed ethnicity: his father was Eritrean and his mother was Ethiopian. The Appellant is married to an Ethiopian citizen. They have a daughter. It seems that the Appellant’s wife and daughter are currently living in Eritrea. The Appellant arrived in the United Kingdom on 6 April 2004 and claimed asylum which was refused on 2 June 2004. The Appellant was consequently refused leave to enter. That decision was, however, subsequently withdrawn. On 12 December 2005, the Appellant was convicted of sexual assault on a female contrary to s.3(1) of the Sexual Offences Act 2003. He was sentenced to 21 months imprisonment and required to sign the Sex Offenders Register for a period of 10 years. The Trial Judge recommended that the Appellant be deported. On 30 June 2006, the Respondent decided to make a deportation order against the Appellant under s.3(6) of the Immigration Act 1971 which (by amendment on 4 October 2006) specified removal to Eritrea or Ethiopia.

2. Although it was accepted that the Appellant had been in the Eritrean military, the Respondent rejected the Appellant’s account of what he claimed had happened to him whilst in the military which, it was said, gave rise to a risk of persecution or serious ill-treatment if he returned. In addition, on the basis of the Appellant’s conviction in the UK under s.3(1) of the Sexual Offences Act 2003, the Respondent concluded that the Appellant, even if a refugee, could nevertheless be returned by virtue of Art 33(2) of the Refugee Convention which provides that the non-refoulement provision in Art 33(1) does not apply if the individual has been convicted of a “particularly serious crime” and he “constitutes a danger to the community”. Applying s.72 of the Nationality, Immigration and Asylum Act 2002 (hereafter “the 2002 Act”) and the Nationality, Immigration and Asylum Act 2002 (Specification of Particularly Serious Crimes) Order (SI 2004/1910) (hereafter “the 2004 Order”), the Respondent concluded that the Appellant was presumed firstly to have been convicted of a ‘particularly serious crime’ and secondly that he constituted a danger to the community which had not been rebutted on the evidence. The Respondent issued a certificate under s.72 of the 2002 Act that the presumptions in s.72(4) of that Act applied.

3. The Appellant appealed against the Respondent’s decision to deport him on a number of grounds. He claimed to be a refugee (alternatively to be entitled to humanitarian protection). He also relied upon Article 3 of the ECHR and para 364 of HC 395. It was argued that Art 33(2) did not apply to the Appellant and that the 2004 Order was ultra vires as there was no power to make it under s.72 of the 2002 Act because it was incompatible with Art 33(2).

4. Following a hearing on 14 December 2006, the Tribunal (Immigration Judge P J M Hollingworth and Dr Chaudhry) (hereafter “the Panel”) dismissed the Appellant’s appeal on all grounds. The Panel found the Appellant not to be credible and rejected his account of what he claimed happened during his military service in its entirety. The Panel did not accept that in 1998, whilst in the military, he was accused of being a spy and had been detained for 6 months in prison and suffered injury as a result of ill-treatment because he had criticised senior officers at a military meeting. The Panel rejected the Appellant’s claim that in February 1999 he was released and rejoined the fighting at the front. Also, the Panel rejected his claim that in April 2000 he was again detained in prison after he had questioned his wife’s deportation to Ethiopia but that 11 days later he escaped after the guards fled when the area was captured by Ethiopian troops. Further the Panel rejected the Appellant’s account that, following his escape, he travelled the 120-130 kilometres to Sudan by foot in 3 days where he stayed between May 2000 and April 2004 when he came to the UK. The Panel also concluded that the Appellant had failed to establish that he had left Eritrea illegally and thus he was not at risk on return on that basis. For those reasons, the Panel rejected the Appellant’s claim under Article 3 of the ECHR. In relation to the Appellant’s reliance upon the Refugee Convention, the Panel agreed with the Secretary of State’s certificate under s.72 that the presumptions in subsection (4) applied and thus dismissed the appeal on refugee grounds. The Panel concluded that s.72 and the 2004 Order (which it held it had no jurisdiction to decide was ultra vires) applied to the Appellant and that therefore the non-refoulement obligation in Art 33(1) of the Refugee Convention did not apply because the Appellant fell within Art 33(2).

5. The Appellant sought, an order for reconsideration, which was granted by Senior Immigration Judge Goldstein on 11 May 2007.

Outline of the Issues

6. We heard oral argument over 3 days in February and September 2008. The parties also put before us extensive skeleton arguments and written submissions, the final one of which was received by the Tribunal on 5 November 2008. We are grateful to the parties for the care and detail in which they presented their submissions.

7. Mr Draycott, who represented the Appellant, challenged the Panel’s decision to dismiss the appeal on the refugee ground and that the Appellant’s removal would not breach Art 3 of the ECHR. We did not understand Mr Draycott directly to challenge the Panel’s decision in respect of para 364 of the Immigration Rules. We do not consider separately in any detail the Appellant’s claim for humanitarian protection because that will stand or fall with our decision in respect of the refugee claim. Mr Draycott made two principal arguments in support of his contention that the Panel had materially erred in law in dismissing the appeal.

8. First, Mr Draycott submitted that the Panel was wrong in law to apply the 2004 Order to the Appellant. The 2004 Order is ultra vires, there being no power in s.72 of the 2002 Act to make an Order inconsistent with Art 33(2) of the Refugee Convention. Because the combined effect of s.72 and the 2004 Order is that certain offences are irrebutably presumed to be ‘particularly serious crime[s]’, the 2004 Order (and s.72 itself) is incompatible with the ‘autonomous’ international meaning of Art 33(2). The enabling power in s.72 has to be read subject to the requirement that it will be exercised consistently with Art 33(2). He submitted that the Tribunal has jurisdiction to decide that the 2004 Order is ultra vires. Further, Mr Draycott submitted that the 2004 Order (and perhaps even s.72 of the 2002 Act itself) could not stand in the face of Art 21 of Council Directive 2004/83/EEC “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” (hereafter the “Qualification Directive”) which imposes on all EU Member States a non-refoulement obligation in the exact terms of Art 33(2) and so through EU legislation gives effect to the ‘autonomous’ meaning of Art 33(2).

9. Secondly, Mr Draycott submitted that the Panel’s decision in respect of Art 3 of the ECHR should be set aside. The Panel erred in reaching its adverse credibility finding that led it to reject the Appellant’s account of what happened to him in, and which caused him to leave, Eritrea. Mr Draycott subjected the Panel’s reasoning on credibility to a detailed critique which, he submitted, showed that the Panel had materially erred in law.

10. Mr Patel, who represented the Respondent, essentially took a contrary stance on each of these issues. His principal submission was that there is no international ‘autonomous’ meaning to be given to Art 33(2) which the Tribunal could apply. The proper meaning to be given to the phrase “particularly serious crime” is a matter for the domestic law of each Contracting State which meant in the UK the legislative provisions in s.72 of the 2002 Act and the 2004 Order. In any event, even if that were wrong, the Tribunal has to apply the regime in s.72 and the 2004 Order as it has no power to hold the 2004 Order to be ultra vires. In relation to credibility, Mr Patel submitted that the Panel’s decision should stand.

11. We deal with the issues raised in this appeal under the following headings: (I) The Legal Framework; (II) Article 33(2) of the Refugee Convention; (III) Compatibility of the 2004 Order and s.72 with Art 33(2); (IV) Ultra Vires and Incorporation; (V) Art 21 of the Qualification Directive; (VI) Credibility.

The Legal Framework

1. The Refugee Convention

12. The relevant provision in the Refugee Convention dealing with refoulement is Art 33 which provides as follows:

“1. No Contracting State shall expel or return (‘refouler’) a refugee in any manner whatsoever to the frontiers of territories where his life or freedom would be threatened on account of his race, religion, nationality, membership of a particular social group or political opinion.

2. The benefit of the present provision may not, however, 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 judgement of a particularly serious crime, constitutes a danger to the community of that country.”

13. Art 33 applies to any individual who establishes that he is a refugee as defined in Art 1A(2) of the Refugee Convention (R v Secretary of State for the Home Department, ex p Sivikumaran [1988] AC 958). Article 33(1) sets out the basic obligation of a Contracting State not to refoule (return) a refugee to his own country where his life or freedom would be at risk for one or more of the so-called five ‘Convention reasons’. That obligation is not, however, without exception. Art 33(2) disapplies that obligation where, inter alia, the refugee has been convicted of a “particularly serious crime” and, as a result, “constitutes a danger to the community” of the country of refuge. It was not disputed before us that the Secretary of State bears the burden of proof in respect of matters relied upon to bring an individual within Art 33(2).

14. The syntax of Art 33(2) is a little tortuous. It creates two different grounds for an exception to the Art 33(1) obligation: (a) “danger to the security” of the country; and (b) conviction of a “particularly serious crime” and a resultant “danger to the community”. At first blush it might be thought that for Art 33(2) to bite there has to be established “reasonable grounds” to regard an individual as a “danger to the community”. But, in fact, the structure of Art 33(2) only engages that phrase (“reasonable grounds”) in cases where the individual is said to be a “danger to the security” of the UK. Otherwise – as in a situation such as arises in this appeal - it must be established that the individual was in fact convicted of the “particularly serious crime” and that he is in fact a “danger to the community”; reasonable grounds alone for so concluding will not suffice (SSHD v TB (Jamaica) [2008] EWCA Civ 977, at [38] per Stanley Burnton LJ).

2. The Qualification Directive

15. The wording of Art 33 is in all material respects reproduced in Article 21.2 of the Qualification Directive. This is important to the Appellant’s argument because it is accepted for the purposes of this appeal by both parties that Art 21 has direct effect and thus, to the extent that it is inconsistent with s.72 or the 2004 Order, it takes precedence over any inconsistent domestic legal provision. Because of the legal effect of EU law, this would avoid any difficulties faced by the Appellant in making good the argument that the Tribunal has jurisdiction to hold the 2004 Order to be ultra vires. It would simply be unlawful to seek to remove the Appellant if that would be contrary to Art 21 of the Qualification Directive and to the extent possible, the UK domestic law should be interpreted to avoid such a conflict. Article 21 is in these terms:

“Protection from refoulement

1. Member States shall respect the principle of non-refoulement in accordance with their international obligations.

2. Where not prohibited by the international obligations mentioned in paragraph 1, Member States may refoule a refugee, whether formally recognised or not, when:

(a) there are reasonable grounds for considering him or her as a danger to the security of the Member State in which he or she is present; or
(b) he or she, having been convicted by a final judgement of a particularly serious crime, constitutes a danger to the community of that Member State.

3. Member States may revoke, end or refuse to renew or to grant the residence permit of (or to) a refugee to whom paragraph 2 applies.”

16. A similar provision is contained in Art 14.4 of the Qualification Directive allowing Member States to revoke, end or refuse to renew refugee status. (Art 14.5 also applies a similar provision where the individual’s refugee status has not yet been recognised.) Mr Draycott placed some reliance upon Art 14 but, in truth, it adds nothing of substance to the Appellant’s case which, in this regard at least, is founded squarely upon the non-refoulement provision in Art 21.2. For this reason we need say no more about Art 14.4.

17. Mr Draycott placed reliance upon Art 21.2 (but not Art 21.1) of the Qualification Directive. He submitted that the ‘non-refoulement’ provision in Art 21.1 is not intended to reflect Art 33 of the Refugee Convention. It was accepted by the parties that the reference to “international obligations” in Art 21.1 is not a reference to the Refugee Convention but rather to such other obligations binding upon the particular state preventing removal or return of the individual, for example in the case of the UK Art 3 of the ECHR or, as suggested by Mr Draycott, Art 3 of the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (1990) to which the UK is a party. That interpretation is borne out in the contrast made in Art 20.6 and 20.7 of the Qualification Directive where reference is made, respectively, to “limits set out by the Geneva Convention” and “limits set out by international obligations”. It is, however, accepted by the parties that Art 21.2 is intended to reflect Art 33 and, in particular in Art 21.2(b), the relevant part of Art 33(2) which is relied upon by the Respondent to permit the Appellant’s refoulement. Where the parties differ is as to the meaning – and thus the effect upon the Appellant - of Art 33(2) which is thereby incorporated into EU law.

3. The Immigration Rules

18. We should also, albeit briefly, refer to the Immigration Rules, Statement of Changes in Immigration Rules (HC 395). Paragraph 334 sets out the criteria for the grant of asylum status. The negative criterion in para 334(iv) reflects the situation where Art 33(1) does, but Art 33(2) does not, apply:

“334. An asylum applicant will be granted asylum in the Unted Kingdom if the Secretary of State is satisfied that: ….

(iv) …having been convicted by a final judgment of a particularly serious crime, he does not constitute [sic] danger to the community of the United Kingdom.”

19. Paragraph 339A(x) of the Immigration Rules sets out an equivalent provision for revocation of, or refusal to renew, refugee status.

4. Section 72 of the 2002 Act

20. We now turn to s.72 of the Nationality, Immigration and Asylum Act 2002, which came into effect on 10 February 2003. Section 72(1) provides that:

“(1) This section applies for the purpose of the construction and application of Article 33(2) of the Refugee Convention (exclusion from protection).”

21. Leaving aside the arguably mistaken reference to “exclusion from protection”, which is more apt to describe the exclusion provisions of Art 1F of the Convention, s.72 provides the UK domestic law means for interpreting and applying the exception to the non-refoulement obligation in Art 33(2) of the Refugee Convention. So far as relevant s.72 provides as follows:

“72. Serious Criminal

(2) A person shall be presumed to have been convicted by a final judgment of a particularly serious crime and to constitute a danger to the community of the United Kingdom if he is -

(a) convicted in the United Kingdom of an offence, and
(b) sentenced to a period of imprisonment of at least two years.

(3) A person shall be presumed to have been convicted by a final judgment of a particularly serious crime and to constitute a danger to the community of the United Kingdom if –

(a) he is convicted outside the United Kingdom of an offence,
(b) he is sentenced to a period of imprisonment of at least two years, and
(c) he could have been sentenced to a period of imprisonment of at least two years had his conviction been a conviction in the United Kingdom of a similar offence.

(4) A person shall be presumed to have been convicted by a final judgment of a particularly serious crime and to constitute a danger to the community of the United Kingdom if –

(a) he is convicted of an offence specified by order of the Secretary of State, or
(b) he is convicted outside the United Kingdom of an offence and the Secretary of State certifies that in his opinion the offence is similar to an offence specified by order under paragraph (a).

(5) An order under subsection (4) –

(a) must be made by statutory instrument, and
(b) shall be subject to annulment in pursuance of a resolution of either House of Parliament.

(6) A presumption under subsection (2), (3) or (4) that a person constitutes a danger to the community is rebuttable by that person.

(7) A presumption under subsection (2), (3) or (4) does not apply while an appeal against conviction or sentence –

(a) is pending, or
(b) could be brought (disregarding the possibility of an appeal out of time with leave).

(8) Section 34(1) of the Anti-terrorism, Crime and Security Act 2001 (c. 24) (no need to consider gravity of fear or threat of persecution) applies for the purpose of considering whether a presumption mentioned in subsection (6) has been rebutted as it applies for the purpose of considering whether Article 33(2) of the Refugee Convention applies.

(9) Subsection (10) applies where -

(a) a person appeals under section 82, 83 , 83A or 101 of this Act or under section 2 of the Special Immigration Appeals Commission Act 1997 (c. 68) wholly or partly on the ground that to remove him from or to require him to leave the United Kingdom would breach the United Kingdom’s obligations under the Refugee Convention, and
(b) the Secretary of State issues a certificate that presumptions under subsection (2), (3) or (4) apply to the person (subject to rebuttal).

(10) The Tribunal or Commission hearing the appeal –

(a) must begin substantive deliberation on the appeal by considering the certificate, and
(b) if in agreement that presumptions under subsection (2), (3) or (4) apply (having given the appellant an opportunity for rebuttal) must dismiss the appeal in so far as it relies on the ground specified in subsection (9) (a).”

22. Importantly for our purposes, ss.72(2) and (4) provide that an individual shall be presumed (i) to have been convicted of a “particularly serious crime”; and (ii) to be a danger to the community of the UK if he is convicted of an offence in the UK and sentenced to at least 2 years’ imprisonment (s.72(2)) or is convicted of an offence specified by order of the Secretary of State (s.72(4)). (Section 72(3) deals with convictions abroad and is not relevant here.)

23. Pursuant to s.72(5), the Secretary of State made the 2004 Order which came into force on 12 August 2004. Schedules 1 to 6 to the Order specify a substantial number of criminal offences in England and Wales, Northern Ireland and Scotland for the purposes of s.72(4). The range of offences is vast covering, for example, crimes relating to the use and possession of chemical, biological or nuclear weapons, involving controlled drugs, crimes of dishonesty, criminal damage, offences against the person, public order offences, firearms offences and sexual offences. The latter includes in Sched 2, para 2 the offence of sexual assault in s.3(1) of the Sexual Offences Act 2003 which is, of course, the crime of which the Appellant was convicted.

24. Section 72(6) provides that the presumption that a person constitutes a danger to the community is rebuttable. Section 72 does not, however, appear on its face to allow a person to rebut the presumption that an offence for which he was convicted and sentenced to at least 2 years’ imprisonment or which is contained in an Order made by statutory instrument under s.72(5) is a “particularly serious crime”. This was assumed to be the position by the Court of Appeal in SSHD v TB (Jamaica) [2008] EWCA Civ 977, at [37] per Stanley Burnton LJ. Mr Draycott relies upon the irrebuttable nature of this presumption as a central part of his argument that the 2004 Order is ultra vires.

25. The presumptions in s.72 apply in all cases, where applicable, heard by the Tribunal (SSHD v TB (Jamaica) [2008] EWCA Civ 977, at [29] per Stanley Burnton LJ). However, s.72(9) and (10) create a mechanism that may be invoked by the Secretary of State which affects the Tribunal’s process when deciding asylum appeals. Section 72(9) permits the Secretary of State to issue a certificate that the presumptions under the relevant subsection of s.72 apply. When this is done, as it was in this case, s.72(10) requires the Tribunal to determine whether the presumptions do in fact apply to the asylum appeal and, if they do, dismiss the appeal. The effect is to alter the normal (and perhaps more natural) way of dealing with a case: first, deciding whether the individual is a refugee; and secondly, only if he is, deciding whether notwithstanding that he may still be refouled because Art 33(2) applies. In cases where a certificate is issued under s.72(9), the refoulement issue is decided first and only if it does not apply can the Tribunal consider whether the individual is in fact a refugee because he has a well-founded fear of persecution for a Convention reason. Of course, in practice the evidence and any risk on return would need to be considered in order to decide whether the individual’s return would breach Art 3 of the ECHR. The decision of the Panel in this appeal is one such example. Section 72(10) does not prevent this, but does prevent the Tribunal from making a finding that he is a refugee albeit one who can be refouled.

Article 33(2) of the Refugee Convention

26. Mr Draycott submitted that there is an ‘autonomous’ or international meaning that must be ascribed to the phrase “particularly serious crime” in Art 33(2) with which the regime in s.72 and the 2004 Order is inconsistent. That submission is crucial to the Appellant’s case before us. If that is not so, there is no valid basis for challenging the Panel’s application of Art 33(2) through the interpretative lens of s.72 and the 2004 Order. It was not contended in the Grounds for Review or before us that the Panel were wrong to conclude on the evidence that the Appellant was a “danger to the community”.

1. Autonomous Meaning?

27. Mr Patel’s submission on behalf of the Respondent is a straightforward one: the Appellant’s challenge cannot succeed because there is no autonomous international meaning which the Tribunal can apply to the phrase “particularly serious crime” in Art 33(2). The interpretation of Art 33(2) and its application is, Mr Patel submitted, a matter for the Contracting States through the mechanism of their domestic law. That argument is, with the greatest respect, hopeless. It might have been possible to argue that an individual phrase, such as ‘particularly serious crime’, should not be taken in isolation from its context, and that the search should be only for the autonomous international meaning of the (whole of the) treaty provision under examination. But Mr Patel did not argue that, and, as will be apparent from the authorities to be cited shortly, courts charged with the interpretation of Art 33(2) have indeed taken this phrase separately and have contrasted its effect with that of other phrases in the paragraph. It follows that even that more refined argument would have been equally hopeless in respect of the particular words with which we are concerned.

28. The Refugee Convention is a treaty binding in international law between the Contracting Parties. (We deal later with Mr Draycott’s argument that it has also been incorporated into our domestic law or forms part of EU law.) As a consequence, Art 33(2) must be interpreted in good faith in accordance with the ordinary meaning to be given to its terms in their context and in the light of its object and purpose (see Art 31(1) of the Vienna Convention on the Law of Treaties (1980) (Cmnd 7964)).

29. In so doing, the domestic courts of a Contracting State, including the Tribunal, must strive to give Art 33 of the Refugee Convention its independent and autonomous international meaning. The point is indisputable and can be illustrated by the House of Lords’ decision in R v SSHD ex p Adan and Aitseguer [2001] 2 AC 477. The House of Lords was concerned with the proper interpretation of Art 1A(2) of the Refugee Convention and whether its protection extended to those who feared persecution by third parties, and not by the State, in their own country or who feared indirect refoulement, i.e. return by another ‘safe’ country to face persecution in their own country. In holding (by a majority of 4-1) that these cases fell within Art 1A(2), the House of Lords sought and applied what Lord Steyn called the “one true interpretation” of Article 1A(2) of the Refugee Convention (at p.518). In his speech Lord Steyn explained the correct interpretative approach as follows (at p.515):
“…the enquiry must be into the meaning of the Refugee Convention approached as an international instrument created by the agreement of contracting states as opposed to regulatory regimes established by national institutions. It is necessary to determine the autonomous meaning of the relevant treaty provision. This principle is part of the very alphabet of customary international law.”
And then (at p.517):
“It follows that, as in the case of other multilateral treaties, the Refugee Convention must be given an independent meaning derivable from the sources mentioned in arts 31 and 32 and without taking colour from distinctive features of the legal system of any individual contracting state. In principle therefore there can only be one true interpretation of a treaty. If there is disagreement on the meaning of the Refugee Convention, it can be resolved by the International Court of Justice: art 38. It has, however, never been asked to make such a ruling. The prospect of a reference to the International Court of Justice is remote. In practice it is left to national courts, faced with a material disagreement on an issue of interpretation, to resolve it. But in doing so it must search, untrammelled by notions of its national legal culture, for the true autonomous and international meaning of the treaty. And there can only be one true meaning.”

30. That approach was accepted by the other Law Lords in the majority (see Lord Slynn of Hadley at p.509, Lord Hobhouse of Woodborough at pp.529-30 and Lord Scott of Foscote at p.531). Of course, it remains a question, to which we shall return, what impact, if any, such autonomous meaning, if inconsistent with the regime in s.72 and the 2004 Order, can have on the UK’s statutory regime. For the present, it suffices to recognise the interpretative task which must be undertaken in this appeal. For these reasons, we reject Mr Patel’s submission that Art 33(2) should not be afforded an autonomous meaning. On the contrary, we must strive to find that meaning in accordance with the relevant principles of interpretation. We cannot give Art 33(2) a purely UK-centred interpretation any more than the House of Lords in Adan and Aitsiguer could give Art 1A(2) such a domestic interpretation.

31. We would, however, make two points. First, we agree with Mr Patel’s submission to this extent: in the absence of a binding international ruling on the meaning of “particularly serious crime”, it is for the domestic courts of each country to seek and apply the ‘autonomous’ meaning of Art 33(2). But in doing so, as the House of Lords made plain in Adan and Aitsiguer, the domestic courts are seeking the “one true meaning”. That was also, it seems to us, the underlying approach (albeit less explicitly articulated) adopted by the House of Lords in T v Immigration Officer [1996] AC 742 when determining the meaning of “serious non-political crime” in Art 1F(b) of the Convention. Secondly, it may be that there is no definitive international definition of particular offences that fall within Art 33(2). After all, unless the substantive international criminal law is to define what is a “crime” for the purposes of Art 33(2) – and it has not been suggested to us that Art 33(2) is so limited – at least that part of the phrase “particularly serious crime” must involve recourse to the domestic criminal law of, presumptively one would anticipate, the country of refuge. Likewise in T, a majority of the Law Lords rejected the view (accepted by Lord Mustill in T) that the word “crime” in the Convention phrase “serious non-political crime” in Art 1F(b) envisaged conduct “recognised as criminal by the common consent of nations”. It may well be, therefore, that the ‘autonomous’ meaning of Art 33(2) entails a wider recourse to the domestic law of the relevant Contacting State in defining what is a “particularly serious crime” than was necessary in the case in Adan and Aitsiguer. We will return to this argument below.
32. Mr Draycott referred us to a number of sources which he relied upon to establish the autonomous meaning of Art 33(2) which, he submitted, is inconsistent with that in s.72 and the 2004 Order, and is more favourable to the Appellant. We consider these materials under the following five headings (a) UNHCR Comments; (b) Parliamentary Joint Committee on Human Rights; (c) Academic Literature; (d) Overseas Case Law; (e) Domestic Case Law.
UNHCR
33. Mr Draycott relied upon the UNHR’s paper, The Nationality, Immigration and Asylum Act 2002 (Specification of Particularly Serious Crimes) Order 2004, UNHCR Comments (November 2004). At page 2 of the document, the UNHCR comment on the application of Art 33(2):
“Article 33(2) applies to refugees who become an extremely serious threat to the country of asylum due to the severity of crimes perpetrated by them in the country of asylum. It aims to protect the safety of the country of refuge and hinges on the assessment that the refugee in question poses a major actual or future threat. For this reason the Article 33(2) mechanism has always been considered as a measure of last resort, taking precedence over and above the application of criminal sanctions and justified by the exceptional threat posed by the individual – a threat such that it can only be countered by removing the person from the country of asylum, including, if necessary, to the country of origin.”
34. In relation to the definition of “particularly serious crime”, the paper continues at page 3:
“UNHCR recognises that the term “serious crime” may have different connotations in different legal systems. In UNHCR’s understanding, the gravity of the crimes should be judged against international standards, not simply by its categorisation in the host State or the nature of the penalty. Crimes such as petty theft or the possession for personal use of illicit narcotic substances would not meet the threshold of seriousness. Examples of a “serious crime”, inter alia, include murder, rape, arson and armed robbery. Certain other offences could be considered serious if they are accompanied by the use of deadly weapons, involve serious injury to persons, or there is evidence of serious habitual criminal conduct. Factors to be considered include the nature of the act, the actual harm inflicted, the form of procedure used to prosecute the crime, and whether most jurisdictions would consider the act in question as a serious crime. The qualification “particularly serious” indicates that only crimes of a particularly serious nature should be considered egregious enough to warrant an exception to the non-refoulement principle.”
35. At page 4, the UNHCR paper continues:
“…it is generally understood that a ‘serious crime’ is a capital or a very grave crime normally punished with long imprisonment….”
36. Applying this, the UNHCR paper concludes that the range of offences contemplated by s.72 and the 2004 Order as constituting “particularly serious crime[s]” is incompatible with the Art 33(2) of the Refugee Convention as representing “a particularly low threshold for an exception to refoulement to apply”.
37. The UNHCR paper also takes issue with the presumptions enshrined in s.72 that a particular crime is a “particularly serious crime” and on the basis of that conviction an individual is a “danger to the community”. These provisions, in the UNHCR’s view, are inconsistent with the Refugee Convention’s requirement that refoulement should only follow a consideration of the individual’s circumstances. At pages 4-5 this is said:
“Importance of individual assessment and proportionality
A judgement on the potential danger to the community necessarily requires an examination of the circumstances of the refugee as well as the particulars of the specific offence. Relevant considerations include whether the refugee may be regarded as incorrigible in light of prior convictions for grave offences, and the prospects for the refugee’s reform, rehabilitation and reintegration into society. Where the refugee has responded to rehabilitative measures, or where there are indications that the refugee can be reformed, Article 33(2) should not apply because the potential threat to the community would have been (or could be) removed. Other relevant considerations would include the refugee’s behaviour while serving his earlier sentence, the fact that they are released on parole, and the refugee’s co-operation in the reform programs.
The particulars of the offence are crucial pointers as to whether the convicted refugee poses or is likely to pose a danger to the community. UNHCR is very concerned more generally that Section 72 NIAA provides for a presumption that an offence is deemed “particularly serious” exclusively on the basis of a custodial sentence of two years or more. This is not appropriate inasmuch as it completely excludes from judicial consideration the overall context of the offence, including its nature, effects and surrounding circumstances, the offender’s motives and state of mind, and the existence of extenuating (or aggravating circumstances). In UNHCR’s view, it is imperative that contextual factors such as these should be considered carefully if Article 33(2) is to be properly applied.
With respect to proportionality, UNHCR has consistently advocated for the need to weigh up the gravity of the offence for which the individual appears to be responsible against the consequences of refoulement. In UNHCR’s view, the proportionality text is necessary in order to ensure that the exceptions applied in manner consistent with the overriding humanitarian object and purpose of the 1951 Convention. Although the concept of proportionality is not expressly mentioned in the 1951 Convention, it is a fundamental principle in international human rights and international humanitarian law (for example, in the jurisprudence of the European Court of Human Rights, as well as in the context of the International Covenant on Civil and Political Rights). UNHCR therefore recommends that proportionality considerations be taken into account when reaching a decision leading to application of Article 33(2).”
38. Whilst much of the UNHCR’s remarks is directed to the “danger to the community” requirement in Art 33(2), there is strong support here for the “contextual” assessment of whether a crime is “particularly serious” based upon a range of factors concerned with the individual and the nature of his past criminal offending.
39. The UNHCR also considers (at p.5) that the shift of the burden of proof brought about by the presumptions in s.72 (particularly in respect of whether the individual is a “danger to the community”) is incompatible with the Refugee Convention although no authority is cited to substantial the proposition:
“Burden of Proof
It is a general principle of law that the burden of proving a particular fact falls upon the party asserting it. In the case of Article 33(2), the burden of proof is on the State to prove that one or several convictions are symptomatic of the criminal, incorrigible nature of the person and that he is likely to do it again, thereby constituting a danger to the community. Both Section 72 NIAA and the list of offences in the Order shift the burden of proof from the State to the refugee to prove that he is not a danger to the community. Such a presumption is, however, excessively onerous and virtually impossible to rebut.”
40. At page 6, the UNHCR reaches its overall conclusion on the UK regime:
“CONCLUSION
In UNHCR’s view, the broad regime set out by the Order and Section 72 NIAA is not necessary or appropriate. UNHCR considers the compilation of a list that removes the need to consider individual cases to be generally undesirable and the long list set out in the Order to be particularly alarming. As outlined above, the analysis to apply Article 33(2) is a two-step process: the first one to establish whether there is a conviction for a particularly serious crime and then whether by this fact, the refugee presents a danger to the community.
In line with general principles of law, the exception to the non-refoulement principles contained in Article 33(2) should be interpreted restrictively. The large number of offences listed in the Serious Crimes Order and the wide discretionary powers granted to the Secretary of State regarding analogous offences committed abroad, instead increase the scope of interpretation. The creation through various pieces of legislation of open-ended exceptions to the rule against non-refoulement, weaken it for all intents and purposes and seriously impair the good-faith application of the 1951 Convention in the United Kingdom.
This is additionally undermined by the use of an administrative instruction extending this wide ranging qualification of particularly serious crimes to deny examination of an asylum claim in substance altogether.”
41. The views of the UNHCR on the scope of Art 33(2) are undoubtedly relevant to our decision. In KK (Article 1F(c)) Turkey) [2004] UKIAT 00101, the IAT set out (at [81]) the proper approach when considering the UNHCR’s view of the meaning of Art 1F(c):
“The views of the UNHCR, which has the responsibility under the governing statute for administering the Refugee Convention as it applies to nations and individuals, are of course entitled to the very greatest respect. Those views are not, however, binding on us and they do not necessarily reflect the correct interpretation of the Convention.”
42. The UNHCR’s paper supports Mr Draycott’s submission that s.72 and the 2004 Order may not reflect the ‘autonomous’ meaning and application of Art 33(2). The need to take account of all the circumstances of the individual and his offending is a central plank of the UNHCR’s understanding of the phrase “particularly serious crime” in Art 33(2). Equally, we see the attraction of an interpretation which is consistent with the humanitarian aims of the Refugee Convention and which, because we are concerned with the scope of an exception to the international protection afforded to refugees by the Convention, militates in favour of an interpretation that only eats into that protection to the extent necessary to protect the country of refuge from a serious criminal who is a danger to that country’s community. That is what we understand by the UNHCR’s reference to interpreting Art 33(2) “restrictively” (see KK at [64] in relation to exclusion from protection by virtue of Art 1F). It is not readily apparent to us, however, why that results in Art 33(2) only being applied in cases where there is an “exceptional threat” to the community. There is no justification of that in the words themselves and we have not been shown any basis for that being the intended scope of Art 33(2).
43. We also have reservations about the UNHCR’s own categorisation of a “serious crime” and “particular serious crime”. No authority or source (beyond the UNHCR’s own Handbook) is offered for the limitation of the latter to a crime that is “capital or very grave … normally punishable with long imprisonment” (p.4). To some extent this cuts across the UNHCR’s own view that the scope and application of Art 33(2) can only be determined by looking at the individual’s particular circumstances. Further the UNHCR offers no basis for the “international standards” which would exclude some crimes from the “serious” category such as possession of narcotics for an individual’s own use but necessarily include, for example, all convictions for murder. As regards the latter, there may be a discernible difference in culpability between a person convicted of murder for helping a terminally-ill loved one to die and a pre-meditated killer of a police-officer on duty. Both are equally guilty of the crime of murder but it may be doubted whether, without more, it would necessarily be right to categorise both as having committed a “particularly serious crime”.
44. The important point, in our view, may well be that made by the UNHCR that circumstances are everything even if some crimes, at first blush, should fall inside or outside the “particularly serious” category. This, of course, is one of Mr Draycott’s points. Section 72 and the 2004 Order do not allow for that individual assessment because the presumptions in s.72(2) (imprisonment for at least 2 years) and s.72(4) (listed offences in the 2004 Order) are irrebuttably treated as “particularly serious” crimes.
Joint Committee on Human Rights
45. Mr Draycott also referred us to the Report of the Parliamentary Joint Committee on Human Rights, The Nationality and Asylum Act 2002 (Specification of Particularly Serious Crimes) Order 2004, Twenty-second Report of Session 2003-04, HL Paper 190, HC 1212 (27 October 2004). The Joint Committee referred to an earlier UNHCR paper, Guidelines on International Protection: Application of the Exclusion clauses: Article 1F of the 1951 Convention relating to the Status of Refugees (4 September 2003), where the UNHCR contrasted the definition of “serious” crime in Art 1F(b) with “particularly serious crime” in Art 33(2):
“25. The Guidelines and Background Note also provide some guidance on the relationship between Articles 1F(b) and 33(2) of the Refugee Convention—
10. Article 1F should not be confused with Article 33(2) … Unlike Article 1F which is concerned with persons who are not eligible for refugee status, Article 33(2) is directed to those who have already been determined to be refugees. Articles 1F and 33(2) are thus distinct legal provisions serving very different purposes. Article 33(2) applies to refugees who become an extremely serious threat to the country of asylum due to the severity of crimes perpetrated by them. It aims to protect the safety of the country of refuge and hinges on the assessment that the refugee in question poses a major actual or future threat. For this reason, Article 33(2) has always been considered as a measure of last resort, taking precedence over and above criminal law sanctions and justified by the exceptional threat posed by the individual—a threat such that it can only be countered by removing the person from the country of asylum.
…
44. Article 1F(b) also requires the crime to have been committed 'outside the country of refuge prior to [the individual's] admission to that country as a refugee'. … Individuals who commit 'serious non-political crimes' within the country of refuge are subject to that country's criminal law process, and in the case of particularly grave crimes to Articles 32 and 33(2) of the 1951 Convention; they do not fall within the scope of the exclusion clause under Article 1F(b). The logic of the Convention is thus that the type of crimes covered by Article 1F(b) committed after admission would be handled through rigorous domestic criminal law enforcement and/or the application of Article 32 and Article 33(2) where necessary.
26. The UNHCR guidance confirms what to us appears from the wording of Article 33(2) compared to Article 1F(b): that the phrase "particularly serious crimes" in Article 33(2) of the Refugee Convention has a narrower meaning than the phrase "serious crimes", which can lead to the exclusion from refugee status under Article 1F(b). Where a refugee who has already been recognised as such commits a serious non-political crime in the country of refuge, this should be dealt with through the ordinary criminal law process of that country; it is only in the case of "particularly grave crimes" that Article 33(2) applies.
Assessment of compatibility
27. We therefore have grave concerns about the compatibility of the Order with the Refugee Convention, properly interpreted. In our view, the crimes included in the Order go far beyond what can be regarded as "particularly serious crimes" for the purposes of Article 33(2). The list of crimes which are specified for the purposes of s. 72(4)(a) of the 2002 Act includes a number of crimes which cannot on any view be regarded as "particularly serious crimes" as that phrase is to be interpreted in the context of the Refugee Convention. It includes, for example, theft, entering a building as a trespasser intending to steal, aggravated taking of a vehicle, criminal damage, and possession of controlled drugs. We doubt whether these offences, per se, would amount to "serious crimes" for the purposes of Article 1F(b), and are even more doubtful that they are capable of amounting to "particularly serious crimes" for the purpose of Article 33(2).
28. By specifying such offences, the Order is in effect expanding the exceptions to the important principle of non-refoulement, and thereby weakening the strength of that principle. In our view this is incompatible with the Refugee Convention properly interpreted. In view of the humanitarian purpose of that Convention, the exceptions to the principle of non-refoulement in Article 33(2) should be given a restrictive interpretation, not an interpretation which expands their scope and correspondingly weakens the principle itself. We draw this matter to the attention of each House.”
46. The Joint Committee also expressed considerable doubt about whether s.72 itself was compatible with the Refugee Convention because of the irrebuttable presumption that certain offences were “particularly serious” and because of the presumption, albeit a rebuttable one, that conviction of such an offence made the individual a danger to the community. At paras 33-36 the Joint Committee said this:
“33. In the course of our consideration of this question, however, it has become clear to us that the very scheme of s. 72 itself may be incompatible with the Refugee Convention in a number of respects. We have concerns, for example, about the appropriateness of a conclusive presumption in relation to whether a crime is "particularly serious". It is clear to us that the applicability of the exceptions to the principle of non-refoulement in Article 33(2) must be judged on a case-by-case basis, taking into account all the individual circumstances of the case, including the particular circumstances of the commission of the specified offence. There may, for example, have been significant mitigating circumstances surrounding the offence which lessen its seriousness. By the same token, there may be aggravating factors which make an otherwise less serious offence more serious, for example the use of a deadly weapon, or serious injury to people. Section 72 precludes any consideration of such factors, because it provides for a conclusive (i.e. non-rebuttable) presumption that certain offences are "particularly serious crimes" for the purposes of Article 33(2).
34. We are also concerned that s. 72 of the 2001 Act reverses the burden of proof in relation to whether a refugee is a "danger to the community". Article 33(2) appears to presuppose that the burden is on the State to establish that an individual, having been convicted of a particularly serious crime, is also a danger to the community. The effect of section 72 of the 2002 Act, however, is that where an offence of a particular kind has been committed by a refugee or a claimant for refugee status, they are presumed both to have committed a particularly serious offence and to be a danger to the community, unless they can demonstrate that they are not a danger to the community.
35. We are further concerned that by adopting a rebuttable presumption approach to the applicability of Article 33(2) of the Refugee Convention, s. 72 of the 2002 Act precludes the application of a proper proportionality test to the particular circumstances of an individual case. In determining whether the exception to the principle of non-refoulement in Article 33(2) applies in a particular case, it is necessary for a balancing exercise to be carried out, weighing the nature of the offence and the degree of danger to the community on the one hand against the degree of persecution feared on the other if the individual were to be returned. Section 72 of the 2002 Act gives effect to an approach to the application of Article 33(2) which has no regard to the individual circumstances of each case other than to the extent that the individual can prove that he is not a danger to the community. Even this limited opportunity to consider the individual circumstances of the case is narrowly circumscribed: the seriousness of the offence is not relevant to this inquiry (see above), and the gravity of the fear or threat of persecution is expressly ruled out as a relevant consideration. The only question is whether the individual can show that he is not a danger to the community. If he cannot prove this, Article 33(2) is deemed to apply simply by virtue of a particular type of offence having been committed.
36. We do not, however, go any further into the question of the compatibility of s. 72 of the 2002 Act with the Refugee Convention, as no purpose would be served reporting to Parliament on the compatibility of a provision which has already been enacted and which there is no imminent occasion for Parliament to reconsider. We merely draw to the attention of each House our concerns about the compatibility of s. 72 of the 2002 Act with Article 33(2) of the Refugee Convention. We may return to this question at an appropriate juncture in the future.”
47. There is considerable overlap between the arguments and conclusions reached by the Joint Committee and those of the UNHCR whose materials the Joint Committee rely so heavily upon. Our earlier concerns about some of the points made by the UNHCR apply with equal force here. We will return to some of these below. For now, we would observe that the restrictive scope afforded Art 33(2) finds no root in the words of Art 33(2) itself, in particular that it serves to except from the non-refoulement obligation only those that pose an “exceptional threat”. Further, even if the contextual approach is the correct one, it is not clear why the Joint Committee reaches the view that Art 33(2) imposes a ‘proportionality’ test requiring the decision-maker to weigh the risk of persecution in a particular case against the nature of the crime and the risk the individual poses to the community. That approach is eschewed in the UK by s.72(8) reading across the effect of s.34(1) of the Anti-terrorism, Crime and Security Act 2001. It has also not found favour in the courts in different but not wholly unrelated contexts (e.g. T v Immigration Officer (Art 1F(b)) and KK (Art 1F(c))). We see no basis for approaching the interpretation of Art 33(2) any differently. One additional point is worth noting. The Joint Committee considered the presumptions in s.72 only implicated the compatibility of s.72 itself with Art 33(2) rather than the vires of the 2004 Order. That may well be correct. The defect will not be with the Order but in its application through the presumptions in s.72(4). Only an argument based upon the Qualification Directive would entitle us to conclude that s.72 is unlawful.
Academic Literature
48. The relevance of academic commentaries to the work of courts in determining the content of international law is not disputed. Mr Draycott referred us to the work of two leading academic authors in the field of refugee law – Professor Guy Goodwin-Gill of Oxford University and Professor James Hathaway (then of) Michigan Law School – in support of his submissions.
49. In The Refugee in International Law (3rd edn) (G. Goodwin-Gill and J McAdam), 2008 OUP, the learned authors cite the majority’s decision in the Australian Federal Court decision of A v Minister for Immigration and Multicultural Affairs [1999] FCA 227 at para [42] to highlight the dual requirements in Art 33(2) of it being established that the individual has been convicted of a “particularly serious crime” and also that he constitutes “a danger to the community”. The authors continue (at pp.237-8):
“Along with a number of other jurisdictions, Australia has now legislated contrary to this reasoning, by prescribing crimes which are to be regarded as ‘particularly serious’ for the purposes of article 33(2) of the 1951 Convention. An approach in terms of the penalty imposed alone will likely be arbitrary. In our view, and as a matter of international law, the interpretation and application of this concept in the context of an exception to non-refoulement ought necessarily to involve an assessment of all the circumstances, including the nature of the offence, the background to its commission, the behaviour of the individual, and the actual terms of any sentence imposed. As in the case of article 1F(b), a priori determinations of seriousness by way of legislative labelling or other measures substituting executive determinations for judicial (and judicious) assessments are inconsistent with the international standard which is required to be applied, and with the humanitarian intent of the Convention. After all, what is at issue here is action by the State in manifest disregard of what is recognized as serious danger (persecution) to the life or liberty of a refugee. It is the nature of presumptions that they disregard context and circumstances, and therefore also the principle of individual assessment.”
50. It is interesting to note from the footnote references (see in particular footnote 236 which we have omitted) that the regime in s.72 and the 2004 Order is not unique to the United Kingdom. Similar legislative “labelling” provisions deeming certain offences to be a “particularly serious crime” exist, for example, in Canada (Immigration and Refugee Protection Act 2001, s.36), Australia (Migration Act 1958, s.91U) and the United States of America (Immigration & Nationality Act 1952, ss.208 and 241).
51. The thrust of Goodwin-Gill and McAdam’s view is that Art 33(2) requires an individual assessment at least of the individual’s danger to the community. Little is said about any ‘autonomous’ meaning of the phrase “particularly serious crime” other than this (at p.237):
“The jurisprudence is relatively sparse and the notion of a ‘particularly serious crime’ is not a term of art…”
52. The learned authors’ views are largely in line with the UNHCR and Joint Committee in relation to the issue of the use of presumptions in s.72. What little is said about the meaning of “particularly serious crime” is, in our respectful view, not helpful. It is unclear what is meant by their latter statement but, in our view, it cannot be taken to mean that there is no autonomous international meaning.
53. In The Rights of Refugees under International Law (2005), CUP Professor James Hathaway offers a fuller analysis of Art 33(2) (at pp.349-52):
“…the criminality exclusion set by Art. 33(2) exists to enable host states to protect the safety of their own communities from criminal refugees who are shown to be dangerous. This right to engage in the refoulement of dangerous criminals is, however, carefully constrained.
First, the gravity of criminality which justifies refoulement under Art. 33(2) is higher than that which justifies the exclusion of fugitives from justice under Art. 1(F)(b) of the Convention. Art. 1 denies protection to an extraditable criminal who has committed a “serious non-political crime outside the country of refuge prior to his admission to that country as a refugee.” “Serious” criminality in this context is normally understood to mean acts that involve violence against persons, such as homicide, rape, child molesting, wounding, arson, drugs trafficking, and armed robbery. The gravity of harm necessary to justify the refoulement of a person who qualifies for refugee status – expressly framed as a “particularly” serious crime – is clearly higher still, and has been interpreted to require that even when the refugee has committed a serious crime, refoulement is only warranted when account has been taken of all mitigating and other circumstances surrounding commission of the offence. ....
Second, while refugee status is to be withheld from persons reasonably suspected of criminal conduct under Art. 1(F)(b), the refoulement of refugees is permissible only when there has actually been conviction by a final judgement. Appeal rights should therefore have expired or been exhausted, limiting the risk of refoulement strictly to those whose criminality has been definitively established in accordance with accepted, general legal norms.
Third and most important, the nature of the conviction and other circumstances must justify the conclusion that the refugee in fact constitutes a danger to the community from which protection is sought. Because danger follows from the refugee’s criminal character, it does not matter whether the crime was committed in the state of origin, an intermediate state, or the asylum state. Nor is it relevant whether the claimant has or has not served a penal sentence or otherwise been punished. In contrast to exclusion from refugee status under Article 1(F)(b) of the Convention, however, particularized refoulement cannot be based on the refugee’s criminal record per se – as seems increasingly to be the practice in the United States, for example. Refoulement is instead authorized only as a “last resort” where there is no alternative mechanism to protect the community in the country of asylum from an unacceptably high risk of harm. The practice of some states to give dangerous refugees the option of indefinite incarceration in the asylum state as an alternative to refoulement is therefore one mechanism to be considered, since it protects the host community, yet averts the risk of being persecuted. In the end, however, the Refugee Convention accepts that in extreme and genuinely exceptional cases, the usual considerations of humanity must yield to the critical security interests of the receiving state. Thus, if the demanding criteria of Art. 33(2) are satisfied , an asylum state may, assuming there is no other option, remove a refugee convicted of a particularly serious crime who poses a danger to the host community’s safety – even if the only option is to send the refugee to his or her country of origin.”
54. We have quoted this passage at length because Professor Hathaway sets out his view of the underlying purpose behind, and consequent limitations upon, Art 33(2) of the Refugee Convention. He states that “the Refugee Convention accepts that in extreme and genuinely exceptional cases, the usual considerations of humanity must yield to the critical security interests of the receiving state.” Again, we remain unconvinced that any words in the text of Art 33(2) or inherent underlying policy necessarily justify such a high hurdle (“extreme” and “exceptional”) for displacement of the non-refoulement obligation. To paraphrase, there is nothing to suggest that only the exceptionally criminal and extremely dangerous fall within Art 33(2). The underlying theme of a “contextual” approach to applying Art 33(2) is also repeated. Hathaway endorses the need for the individual’s circumstances to be taken into account in assessing the application of Art 33(2). To this extent, Hathaway’s understanding of the meaning of Art 33(2) chimes with the essence of Mr Draycott’s submissions.
55. To support this point, Professor Hathaway relies upon the Australian Federal Court decision of Betkoshabeh v Minister for Immigration and Multicultural Affairs (1998) 157 ALR 95 (reversed on other grounds: [1999] FCA 980). This case is also relied upon by Mr Draycott and we consider it next.
Overseas Case Law
56. Mr Draycott referred us to the Australian Federal Court decision of Betkoshabeh v Minister for Immigration and Multicultural Affairs (1998) 157 ALR 95. In Betkoshabeh, the appellant was an Assyrian Christian from Iran. He was recognised as a refugee in Australia. The appellant developed severe paranoid delusions. He formed the view that an interpreter whom he had befriended whilst detained pending resolution of his refugee status was responsible for his detention and involved in a conspiracy to deport him. He entered the interpreter’s home and, armed with two knives, hid in a cupboard. He was found by the interpreter’s father and subsequently pleased guilty to charges of being unlawfully on the premises and intentionally damaging property. He received a community based order. Subsequently, he went again to her home with a knife and threatened to kill her. He made subsequent threats over the telephone. He was convicted of one count of aggravated burglary and five counts of threats to kill. He was sentenced to three years and six months imprisonment.
57. A decision was made to deport the appellant. The Minister relied upon Art 33(2) and the Administrative Appeals Tribunal affirmed the decision to deport him on the basis that the offence of “threat to kill” was a “particularly serious crime” within Art 33(2) of the Refugee Convention. The appellant appealed to the Federal Court which quashed the decision. Finkelstein J held that the AAT had misconstrued Art 33(2) by failing to consider the specific circumstances of the appellant’s crime (particularly his psychiatric condition) in determining that it was one that fell within Art 33(2). Finkelstein J said (at p.100):
“The expression "particularly serious crime" in Article 33(2) is not defined in the Convention. The expression shows that it is not enough for the crime committed to be a serious crime. It must be "particularly serious" as well as a crime that shows that the refugee is a danger to the community.
On its proper construction, Article 33(2) does not contemplate that a crime will be characterised as particularly serious or not particularly serious merely by reference to the nature of the crime that has been committed although this may suffice in some cases. The reason is that there are very many crimes where it is just not possible to determine whether they are particularly serious without regard to the circumstances surrounding their commission. ”
58. Having cited the earlier unreported case of Vabaza v Minister for Immigration and Multicultural Affairs (Fed Ct Aust, Goldberg J, 27 February 1997) in which the judge seemed to say that the offences of a threat to kill and rape were per se “particularly serious crimes”, Finkelstein J continued (at pp.101-2):
“If this is what his Honour meant then I regret to say that I am unable to agree with him. There will be occasions when a threat to kill cannot be treated as a particularly serious crime. It all depends upon the circumstances. While it is true that rape is a serious crime there will be occasions, rare though they may be, when a rape could not be treated as a particularly serious crime. Again, it all depends upon the circumstances.
The view that I have formed concerning the proper approach to be adopted in determining whether a crime is particularly serious for the purposes of Article 33(2) is one that has been applied in the United States of America. For example, in In the matter of Frentescu, (1982) 18 I & N Dec 244 a refugee had been convicted of burglary and sentenced to a term of imprisonment of three months. The question arose whether this was a conviction for a "particularly serious crime" within the meaning of s 243(h)(2)(B) of the Immigration and Nationality Act 1952 (US) thus enabling the refugee to be deported from the United States. Section 243(h)(2)(B) was in substantially the same terms as Article 33(2) of the Convention. The Board of Immigration Appeals held that while there are crimes which on their face are particularly serious crimes, in most cases it is necessary to analyse each crime on a case by case basis to decide whether it is particularly serious. The Board said (at 247):
"In judging the seriousness of a crime, we look to such factors as the nature of the conviction, the circumstances and underlying facts of the conviction, the type of sentence imposed, and, most importantly, whether the type and circumstance of the crime indicate that the alien will be a danger to the community. Crimes against persons are more likely to be characterised as "particularly serious crimes". Nevertheless we recognise that there may be instances where crimes (or a crime) against property will be considered as such crimes".
This formulation of principle was approved by two decisions of the Court of Appeals for the 9th Circuit, namely Ramirez-Ramos v Immigration and Naturalisation Service (1987) 814 F (2d) 1194 and Beltran-Zavala v Immigration and Naturalisation Service (1990) 912 F (2d) 1027.
In Beltran-Zavala the Court of Appeals was concerned with a refugee who had been convicted of selling marijuana to an undercover police officer. The refugee, Beltran, pleaded guilty to a violation of the Californian Health and Safety Code and was sentenced to two years probation. Shortly thereafter Beltran was arrested for an alleged theft from an automobile. As a consequence he had his probation revoked. While imprisoned Beltran received from the Immigration and Naturalisation Service an order to show cause why he should not be deported. An immigration judge held that Beltran did not qualify for the withholding of deportation because he had been convicted of a "particularly serious crime". An appeal from this decision to the Board of Immigration Appeals was dismissed and that decision was the subject of review by the Court of Appeals. The court reversed the decision of the Board. It said (at 1421):
"In the case at hand, however, the BIA did not examine the type of sentence or the underlying facts. It simply leapt directly from the fact of conviction to the determination that it could not withhold deportation. In fairness, it may not have given as much consideration to this factor as it should have, because of its error regarding asylum. Had it been correct about Beltran's eligibility for asylum, the BIA could have concluded that Beltran was not entitled to withholding of deportation in any event. The BIA was not correct."
In this case the Tribunal fell into the same error. It failed to have regard to the facts and circumstances underlying the commission of the various offences of which the appellant had been convicted. It simply decided that those offences were "particularly serious offences" because of the nature of those offences. The Tribunal should have taken into account the fact that it was the appellant's psychological illness that led to the commission of the offences. It should have taken into account that the appellant's conduct was directed to a person whom he believed, as a consequence of his psychological illness, had been conspiring to cause him harm. The Tribunal should have considered the extent to which that psychological illness reduced the moral culpability of the appellant in much the same way as his psychological illness was taken into account in sentencing the appellant for having committed those offences: as to the relevance of a psychological illness in sentencing ….”
59. Mr Draycott relied upon the decision of Finkelstein J in support of his contention that the autonomous meaning of Art 33(2) was inconsistent with the 2004 Order and the presumptions in s.72.
60. For completeness we should refer to the case of A v Minister for Immigration and Multicultural Affairs [1999] FCA 227. Mr Draycott referred to A in his written submissions and it is discussed in Goodwin-Gill and McAdam (at pp.238-9) to which we were also referred. We were not supplied with a report of the case but we subsequently obtained a transcript of the judgments via BAILII. A number of issues were raised in the case. Dealing with the phrase “particularly serious crime” in Art 33(2), Burchett and Lee JJ (in their majority judgment) approved the approach in Betkoshabeh (at [4]):
“Finkelstein J made it clear that a crime will not necessarily "be characterised as particularly serious or not particularly serious merely by reference to the nature of the crime", although he thought it might be in a particular case. That was because its seriousness would depend on the circumstances surrounding its commission.”
61. Katz J (the third member of the court) also accepted Finkelstein J’s approach and referred to its acceptance in further litigation involving the same claimant (at [41]):
“As to whether A had been convicted of a particularly serious crime within the meaning of Art 33(2), in Betkoshabeh v Minister for Immigration and Multicultural Affairs (1998) 157 ALR 95 at 100 (Finkelstein J) held that, in order to determine whether a crime is a particularly serious one for present purposes, it is generally necessary to have regard to the circumstances in which it was committed, although he accepted the possibility that there can be crimes which are particularly serious per se. In Betkoshabeh v Minister for Immigration and Multicultural Affairs [1999] FCA 16 (unreported, 15 January 1999), a later case involving the same parties, Marshall J (at par 8 of his reasons for judgment) agreed with the approach of Finkelstein J.”
Domestic Case Law
62. As regards this country, there is a dearth of relevant case law. We were only referred to one case touching on the issues we are asked to decide in respect of Art 33(2) namely SB (Cessation and Exclusion) Haiti [2005] UKIAT 00036. The appellant was a refugee who had a string of convictions including ones for assault occasioning actual bodily harm, theft, burglary, possession of an imitation firearm and wounding. In respect of the latter two offences he had been sentenced to terms of imprisonment of 15 months and 3 years respectively. He sought to resist deportation, inter alia, on the basis of Art 33(1). The Secretary of State relied upon Art 33(2). Section 72 was not in force at the relevant time. The IAT rejected the Secretary of State’s submission that a series of offences could amount to “a particularly serious crime” (at [70]):
“The Convention, and section 72, refer to “a” particularly serious crime. Of course, convictions for several “particularly serious crimes” also suffices, but convictions for several “not particularly serious crimes” do not cause the protection against non-refoulement to be removed. There must be at least one conviction for “a particularly serious crime”. This is not a piece of pedantic focus on an indefinite article. The removal of protection is serious; the disqualifying offence has to attain a particular level not met by persistent low level offending. Incorrigible criminality and danger to the community do not suffice of themselves.”
63. In dismissing his appeal, the IAT took the view that on the facts only the offence of wounding was a candidate for being a “particularly serious offence” (at [72]). The IAT looked at the circumstances surrounding its commission (at [73]) and concluded (at [74]):
“74. This offence, by itself or as a single offence, does not quite reach the threshold, even though it was a nasty attack, on people in their homes, and undertaken because the Appellant thought that he was entitled to take the law into his own hands. However, in the context of the Appellant’s previous offending, which involved dishonesty and violence, and an earlier occasion on which he had gone to someone’s house to deal with a perceived wrong, outside the law, it does reach that threshold.”
64. Mr Draycott relied on the IAT’s approach and submitted that it was only the context of the appellant’s cumulative criminality that ‘tipped the balance’ in respect of a conviction for wounding that resulted in a 3 year prison sentence. In the light of this, applying the autonomous meaning of the phrase “particularly serious crime”, it cannot, he submitted, be said that the Appellant’s offence under s.3(1) of the Sexual Offences Act 2003 for which he was sentenced to a term of 21 months reached the high threshold to fall within Art 33(2).
2. Analysis and Conclusions
65. On the basis of the materials to which we have been referred and the submissions made to us, we make the following preliminary points.
66. First, it is our task to search for the autonomous international meaning of the phrase “particularly serious crime” in Art 33(2). That task is to be undertaken in accordance with the principles of interpretation set out in the Vienna Convention, in particular Art 31 which we have set out above. There is no binding decision upon us, either from within the UK or at an international level, as to the proper interpretation of Art 33(2).
67. Secondly, there is no clear guidance in the text of the Refugee Convention itself. There is an apparent contrast drawn in the text between the phrase “particularly serious crime” in Art 33(2) and “serious non-political crime” in Art 1F(b) setting out one of the bases for excluding an individual from refugee status. The addition of the word “particularly” in Art 33(2) suggests a higher threshold than is required for exclusion under Art 1F(b).
68. Given the different applications of Arts 1F and 33, it is not clear to us that any meaningful interpretative assistance beyond this can be gained by dwelling on the linguistic variation. The arguments pull, at times, in conflicting directions. Art 1F(b) deals with exclusion from the Convention’s protection rather than, as does Art 33(2), permitting refoulement of a person otherwise falling within the Convention’s protection. On the one hand, one might have expected the more drastic outcome that an individual is excluded from the Convention’s protection to require a higher, rather than the apparently lower, threshold. That is, perhaps, reflected in the very serious nature of the international crimes covered by Art 1F(a), e.g. crimes against humanity and war crimes, and by the high level of international transgression that is needed to amount to acts “contrary to the purposes and principle of the United Nations” in Art 1F(c). But, it is not apparently reflected in the respective wording of Art 1F(b) and Art 33(2). On the other hand, Art 33(2) is only engaged where an individual is a refugee under the Refugee Convention: subject in the UK to the effect of s.72(10) in a certified case. Art 33(2) does not take away that status. As the IAT remarked in KK at [81]:
“[Articles 32 and 33] do not affect his refugee status: they merely diminish the incidents of that status.”
The consequence is that Art 33(2) contemplates the return of an individual who has established a risk on return to their own country. It might be thought that Art 33(2) potentially has therefore an even greater impact upon an individual which would justify a higher threshold than Art 1F(b) before it is engaged. But even that argument loses some force when it is appreciated that the exclusion provisions in Art 1F can be applied to an individual who has already been recognised as a refugee (see, MT (Algeria) v SSHD [2007] EWCA Civ 808 approving KK at [87]).
69. Thirdly, there is broad agreement in the material to which we have been referred that for Art 33(2) to be engaged all the circumstances of the individual must be considered. For some this is because, usually, whether a crime is “particularly serious” cannot be determined merely by looking at the nature of the crime; the individual circumstances of the commission of the offence must be taken into account. Others more readily tether the contextual approach to the “danger to the community” requirement. By way of possible internal contradiction, there is broad agreement that some offences, for example murder always fall within the “particularly serious” category. Absent legislative provisions, both in the UK and abroad, judges have approved and applied this approach of looking at the individual circumstances of the commission of the offence in order to determine whether it is a “particularly serious crime” within Art 33(2). Of particularly note is the Australian case of Betkoshabeh.
70. Fourthly, and perhaps by way of counter-weight to the previous point, it may not be without significance that in a number of countries who are parties to the Refugee Convention, for example Canada, Australia and the USA, legislation has been enacted to specify that certain crimes will be treated as “particularly serious” for the purposes of Art 33(2) or its legislative equivalent in that country – what Goodwin-Gill and Adams refer to as legislative “labelling”. We were not taken to this material by either representative. We are not, therefore, in a position to analyse it and we can draw no conclusions from its mere existence. It may be that each of these countries has acted contrary to the Treaty obligations or it may be that their actions are indicative of what has been understood that Contracting States may do consistently with their obligations under the Refugee Convention.
71. Where does that leave the proper interpretation (and application) of Art 33(2)?
72. We are in no doubt that what amounts to a “particularly serious crime” is (in part) anchored in the domestic law of the country of refuge (or abroad where that is the place of conviction relied upon). We do not see how the notion of a “crime” can be otherwise defined once international criminal law is (rightly) not considered to be the touch-stone. That does not mean, however, that when a crime is to be characterised as “particularly serious” is also a matter of domestic law. It is not: there is an ‘autonomous’ international meaning of the phrase.
73. That said, no ‘bright line’ definition has been suggested to us and none is offered in the materials to which we were referred. What we find there is, perhaps, the next best thing; namely an interpretative approach that recognises that, in applying Art 33(2) to a specific individual, consideration must be taken of the individual circumstances of the commission of the offence. That, in our view, is the correct interpretation – and interpretative approach to – Art 33(2) and the phrase “particularly serious crime”. We essentially agree with Finkelstein J’s view in Betkoshabeh ( at p.100) that:
“On its proper construction, Art 33(2) does not contemplate that a crime will be characterised as particularly serious or not particularly serious merely by reference to the nature of the crime that has been committed although this may suffice in some cases. The reason is that there are very many crimes where it is just not possible to determine whether they are particularly serious without regard to the circumstances surrounding their commission.”
74. Thus, the forensic investigation of whether a crime is a “particularly serious” one in a given case must be a struggle by the decision-maker (judicial or otherwise) with the facts and circumstances relating to the conviction and the offender. The conclusion, which is part factual but essentially one of judgment, must be conscientiously and rationally arrived at. No definitional or descriptive lexicon has been suggested to us that can assist further in the judicial quest for what is a “particularly serious crime”. Despite its popularity in the non-judicial materials we have been referred to, we see no basis for restricting the application of Art 33(2) to “exceptional” or “extreme” cases or to “grave” crimes. To do so would, in our judgment, be to impose an unwarranted gloss upon the words of Art 33(2).
75. We accept that the words of Art 33(2) dictate that not every crime will suffice – self-evidently the crime must be “particularly serious”. We also agree with the Australian Federal Court in A (discussed more fully below) that it is important to bear in mind that the requirement that the individual is a “danger to the community” is a distinct issue when applying Art 33(2). It may well be relevant to consider the dangerousness of the crime (and of the criminal) in determining whether someone has been convicted of a “particularly serious crime” but it can be no more than a relevant factor. One requirement may be established even if the other is not.
76. We agree with Finkelstein J’s statement in Betkoshabeh that in determining whether an individual’s conviction falls within Art 33(2), it may be that some offences by their very nature will usually be covered. There will be little need to look at the circumstances. Murder, rape, serious crimes of violence or public disorder may require little more to place them in the “particularly serious” category. But that will be because nothing in the circumstances dilutes the justifiable conclusion reached. But not always: it is not difficult to envisage even for murder circumstances which might lead to a contrary view, for example the offence is an act of mercy killing. Likewise, there are crimes which, at first blush, do not necessarily seem to be “particularly serious”, for example theft. Theft is, of course, one of the scheduled crimes in the 2004 Order. By virtue of that Order, it is always a “particularly serious crime” regardless of the circumstances of its commission. But, there may be a world of difference between the categorisation of a conviction for theft arising from a bank heist and one where the individual stole from a supermarket to feed his family. Both amount to the same crime, but only by looking at the circumstances can it properly be determined whether the crime was in fact “particularly serious”. Views may differ on these factual situations and that is the point. In our view, Art 33(2) can only be applied in a fact-sensitive way taking account of all the circumstances of the offence including its nature, gravity and consequences and of the offender including any aggravating or mitigating factors.
Compatibility of the 2004 Order and s.72 with Art 33(2)
77. It follows from our analysis above that by depriving the decision-maker of the ability to take account of all the circumstances of an offence’s commission, and instead listing crimes which a priori fall within Art 33(2), the legislative scheme of the 2004 Order combined with the irrebuttable presumption in s.72(4) that offences falling within it are “particularly serious crime[s]” diverges from the ‘autonomous’ international meaning and interpretation of Art 33(2). (The same conclusion follows ineluctably for the presumptive effects of ss.72(2) and (3) of the 2002 Act in respect of a UK conviction with a sentence of at least two years’ imprisonment and the equivalent overseas conviction.)
78. We do not accept Mr Draycott’s submission that any presumption would offend the Refugee Convention. Despite the contrary view of the UNHCR (at p.5) and the “concerns” of the Joint Committee (at paras 34-35), we see nothing in principle that should lead us to conclude that s.72(4) (or indeed s.72(2) or (3)) offends the ‘autonomous’ meaning and interpretation of Art 33(2) in presuming that a specific crime is a “particularly serious” one providing that presumption is rebuttable. The individual would have an opportunity to raise the relevant individual circumstances albeit that a burden of proof is prima facie cast upon him by the presumption. Only if the presumption could not be rebutted even in the face of evidence and argument to the contrary would the proper effect of Art 33(2) be frustrated. The problem is that, on ordinary principles of interpretation, s.72(4) does create an irrebuttable presumption (see, SSHD v TB (Jamaica) [2008] EWCA Civ 977, at [37] per Stanley Burnton LJ).
79. That deficiency is not, in our view, ameliorated by the fact that there is the additional requirement in Art 33(2) that the individual must be shown to be “a danger to the community”. That requirement is also subject to a presumption against the individual but it is one that is rebuttable by virtue of s.72(6) of the 2002 Act and which might well therefore permit at least some individual, contextual consideration of the relevant circumstances. The two are nevertheless separate requirements in Art 33(2) that must each be satisfied for Art 33(1) not to apply. We agree with the view of Burchett and Lee JJ in the Federal Court of Australia‘s decision in A v MIMA (at [3]-[4])that:
“The logic of the syntax of [Art 33(2)] moves in the opposite direction. The principal statement of exclusion is ‘who constitutes a danger to the community’. The phrase ‘having been convicted…of a particularly serious crime’ adds an additional element, but it is not expressed as if the additional element swallowed up the principal statement….both of two things are required, not that one of them negates the need to consider the other.”
80. Although the Australian court was concerned to reject the converse argument, namely that merely to be convicted of a “particularly serious crime” sufficed to show the person was a “danger to the community”, the interpretative guidance is as apposite here. Merely showing that an individual is a “danger to the community” cannot obviate the need to show that the danger springs from his conviction for a “particularly serious crime”. The latter requires the individual circumstances pertaining to that offence to be considered which may or may not correlate directly to the factors relevant to determining whether he is also a “danger to the community”.
81. That conclusion, however, as Mr Draycott recognised does not assist the Appellant to win this appeal unless the ‘autonomous’ meaning has effect in English law and further that the inconsistency means either that the legislative scheme must be interpreted consistently with the ‘autonomous’ meaning or, if cannot be, the Tribunal is required not to apply it because the 2004 Order is ultra vires and if EU law is in play, s.72(4) itself is unlawful.
Ultra vires and Incorporation
1. Ultra Vires?
82. We deal first with the issue of whether the Tribunal has jurisdiction to determine whether a statutory instrument, here the 2004 Order, is ultra vires and, if so, not apply it.
83. The 2004 Order is a statutory instrument passed subject to the negative resolution procedure in Parliament (s.72(5)). As such, it became law on 12 August 2004. It is trite law that a statutory instrument, unlike an Act of Parliament, may be held to be ultra vires on substantive or procedural grounds. (Statutory instruments are also subject to challenge under the provisions of the Human Rights Act 1998, including the interpretative imperative in s.3(1).) A statutory instrument is, nevertheless, the law unless and until it is held to be ultra vires (usually by declaration) by a competent court: until such time it has legal and practical effect (Hoffman-La Roche (F) & Co v Secretary of State for Trade and Industry [1975] AC 295).
84. The conventional view is that the ‘competent’ court in England and Wales is the High Court through an application for judicial review under Part 54 of the CPR or, in Scotland, is the Court of Session exercising its comparable supervisory jurisdiction. It does not include tribunals such as the AIT. Thus, in a number of Court of Appeal decisions where parts of the AIT’s 2005 Procedural Rules have been challenged as being ultra vires it has never been suggested that the Tribunal could itself decide that the rule was ultra vires rather it has been assumed that the Court of Appeal alone, despite hearing a statutory appeal from the AIT on a point of law, exclusively has that jurisdiction presumably under its inherent jurisdiction: see AM (Serbia) v SSHD [2007] EWCA Civ 19 (r.62(7)); FP (Iran) v SSHD [2007] EWCA Civ 13 (r.19(1) – although these cases may do no more than reflect the fact that an immigration decision cannot be said to be “not in accordance with the law” merely because the AIT applies an ultra vires procedural rule in determining an appeal against that decision after it has been made).
85. Likewise, the Tribunal has itself – with one exception – eschewed such a jurisdiction based upon the appeal ground (now in ss.84(1)(e) and 86(3)(a)) that the decision is “not in accordance with the law” both in respect of statutory instruments and the Immigration Rules which, of course, are not a statutory instrument but merely formalised policies of the Secretary of State albeit ones given statutory imprimatur by s.3(2) of the Immigration 1971 (see Odelola v SSHD [2008] EWCA Civ 308 and AM (Ethiopia) and others v SSHD [2008] EWCA Civ 1082).
86. In Pardeepan v SSHD [2000] INLR 447 Collins J speaking for the IAT stated (at [7]) that:
“…we would need a very great deal of persuasion indeed that we had the power to disregard any legislative provision, be it primary or secondary legislation.”
87. Any reticence on Collins J’s part had disappeared a few months later when in Koprinov v SSHD (01/TH/00091) (5th February 2001), he said:
“We are…satisfied that, although ‘the general law’ and ‘established principles of administrative or common law’ are to be applied, adjudicators and the tribunal cannot decide whether Immigration rules or any subordinate legislation is ultra vires.”
88. The same view also found favour in R v IAT ex parte Begum [1986] Imm AR 385. At page 389 Simon Brown J (as he then was) said this:
“[The submission that the Immigration Rules should be struck down as ultra vires] is not one which could have been advanced before the Tribunal. Only the High Court has power under its supervisory jurisdiction to strike down such a provision. The Tribunal, having construed it, had simply to apply it, however unreasonable and unfair they regard it to be.”
89. Likewise in AI v SSHD [2007] EWCA Civ 386, Dyson LJ at [58] re-iterated the same limited scope to the Tribunal’s jurisdiction in relation to the immigration rules:
“If para 289A(iv) did bear the meaning contended for by the Secretary of State and if it were ultra vires, the AIT should nevertheless have given it full force and effect. They have no power to declare it to be ultra vires, nor did they purport to do so.”
90. That view applies a fortiori in respect of a statutory instrument such as the 2004 Order.
91. The only contrary decision to which we were referred directly dealing with the Tribunal’s jurisdiction is Shafique v SSHD (18448) (16 September 1998) where, relying on the House of Lords’ decision in Foster v Chief Adjudication Officer [1993] AC 794, the IAT declined to follow Simon Brown J’s judgment in Begum and concluded that it had jurisdiction to determine the vires of an immigration rule. At p.10, the IAT said this:

“In the light of the interpretation put upon the phrase “in accordance with the law…applicable” by the Court of Appeal in Abdi (above), whereby an adjudicator or the Tribunal is permitted to consider whether a decision is in accordance with the established principles of administrative or common law, it is hard to see why the issue of vires should be excluded from consideration. In the Tribunal’s view the approach taken by the House of Lords in Foster (above) in respect of the jurisdiction of Social Security Commissioners should apply equally and for the same reasons to the Immigration Appeal Tribunal. In our view the decision of Manshoora Begum (above) should be treated as superseded by Foster (above). Accordingly the Tribunal are satisfied that they do have jurisdiction to consider the vires of the Immigration Rules.”
92. Were it not for the decision in Foster and another decision of the House to which we were referred, Boddington v British Transport Police [1999] 2 AC 143, we would have no hesitation in rejecting Mr Draycott’s submissions and accepting those of Mr Patel to the effect that the Tribunal has no power to disregard a statutory instrument such as the 2004 Order because it considers it to be ultra vires. The Tribunal only has statutory jurisdiction (so far as relevant for these purposes) to determine whether an immigration decision is “not in accordance with the law” (s.84(1)(e) and s.86(3)(b)). That “law” includes any statutory instrument unless and until it is held to be invalid by the High Court. The legal ‘metewand’ by which the Tribunal determines the legality of the ‘immigration decision’ is that “law”. A challenge that entails the argument that the “law” itself is unlawful is a more deep-rooted and fundamental challenge going beyond the legality of the ‘immigration decision’ itself. It is not one which we consider to be contemplated by the 2002 Act. It is properly the domain of judicial review. It is true that in AA and others [2008] UKAIT 00003 the Tribunal took the view that the ground of appeal in s.84(1)(e) permitted challenges based upon public law principles, for example of fairness and legitimate expectation (see especially at [55]). That, however, was intended as a general statement of the scope of the statutory ground: it did not purport to deal with the issue in this appeal. The Tribunal was not concerned with the vires of a statutory instrument or of an immigration rule. Indeed, at para [51] the Tribunal specifically noted that such challenges fell exclusively within the jurisdiction of the Administrative Court (or Court of Session). That is also, in our view, a complete answer to Mr Draycott’s reliance upon the dicta in E v SSHD [2004] 2 WLR 1351 where the Court of Appeal stated that the grounds of challenge in a case on appeal on a point of law were the same as those available on judicial review. The Court simply did not have in mind the issue raised here.
93. Our understanding of the law, however, requires re-examination in the light of the decisions in Foster and Boddington which were relied upon by Mr Draycott. We begin with Boddington.
94. In Boddington, the defendant was convicted in the Magistrates Court of the offence of smoking a cigarette in a railway carriage where smoking was prohibited (by notice) contrary to the British Railway Board’s Byelaws 1965 made pursuant to s.67(1) of the Transport Act 1962. The defendant sought to raise the defence that the relevant byelaw (taken with the required administrative act of posting a notice prohibiting smoking) was ultra vires the 1962 Act. Hence, he had not committed the offence. Overturning the decision of the Divisional Court, the House of Lords concluded that the defendant could challenge the vires of the byelaw (and administrative act) in the criminal proceedings, although the House held that he had not made out the substance of his challenge. The House of Lords held that the defendant was not required to seek judicial review in the High Court. The reasoning of the Law Lords reflects the fact that the defendant faced a criminal charge and the impracticalities of pursuing judicial review in such cases.
95. Lord Steyn commented (at p.226):
“One would expect a defendant in a criminal case, where the liberty of the subject is at stake, to have no lesser rights. Provided that the invalidity of the byelaw is or maybe a defence to the charge a criminal case must be the paradigm of collateral or defensive challenge.”
96. Later in his speech Lord Steyn (at p.227) considered the problems of restricting the defendant’s ability to challenge the vires of the byelaw to judicial review proceedings as had been held in the earlier case of Bugg v DPP [1993] QB 473:
“That decision contemplates that, despite the invalidity of a byelaw and the fact that consistently with Reg. v. Wicks such invalidity may in a given case afford a defence to a charge, a magistrate court may not rule on the defence. Instead the magistrates may convict a defendant under the byelaw and punish him. That is an unacceptable consequence in a democracy based on the rule of law. It is true that Bugg's case allows the defendant to challenge the byelaw in judicial review proceedings. The defendant may, however, be out of time before he becomes aware of the existence of the byelaw. He may lack the resources to defend his interests in two courts. He may not be able to obtain legal aid for an application for leave to apply for judicial review. Leave to apply for judicial review may be refused. At a substantive hearing his scope for demanding examination of witnesses in the Divisional Court may be restricted. He may be denied a remedy on a discretionary basis. The possibility of judicial review will, therefore, in no way compensate him for the loss of the right to defend himself by a defensive challenge to the byelaw in cases where the invalidity of the byelaw might afford him with a defence to the charge. My Lords, with the utmost deference to eminent judges sitting in the Divisional Court I have to say the consequences of Bugg's case are too austere and indeed too authoritarian to be compatible with the traditions of the common law. In Eshugbayi Eleko v. Government of Nigeria [1931] A.C. 662, a habeas corpus case, Lord Atkin observed, at p. 670, that "no member of the executive can interfere with the liberty or property of a British subject except on condition that he can support the legality of his action before a court of justice." There is no reason why a defendant in a criminal trial should be in a worse position. And that seems to me to reflect the true spirit of the common law.
There is no good reason why a defendant in a criminal case should be precluded from arguing that a byelaw is invalid where that could afford him with a defence.”
97. To like effect, Lord Irvine of Lairg LC said (at pp.216-7):
“…where subordinate legislation (e.g. statutory instruments or byelaws) is promulgated which is of a general character in the sense that it is directed to the world at large, the first time an individual may be affected by that legislation is when he is charged with an offence under it: so also where a general provision is brought into effect by an administrative act, as in this case. A smoker might have made his first journey on the line on the same train as Mr. Boddington; have found that there was no carriage free of no smoking sign and have chosen to exercise what he believed to be his right to smoke on the train. Such an individual would have had no sensible opportunity to challenge the validity of the posting of the no smoking signs throughout the train until he was charged, as Mr. Boddington was, under Byelaw 20. In my judgment in such a case the strong presumption must be that Parliament did not intend to deprive the smoker of an opportunity to defend himself in the criminal proceedings by asserting the alleged unlawfulness of the decision to post no smoking notices throughout the train. I can see nothing in section 67 of the Transport Act 1962 or the byelaws which could displace that presumption. It is clear from Wandsworth London Borough Council v. Winder [1985] A.C. 461 and Reg. v. Wicks [1998] A.C. 92, 116, per Lord Hoffmann that the development of a statutorily based procedure for judicial review proceedings does not of itself displace the presumption.
Accordingly, I consider that the Divisional Court was wrong in the present case in ruling that Mr. Boddington was not entitled to raise the legality of the decision to post no smoking notices throughout the train, as a possible defence to the charge against him.”
(Lords Slynn of Hadley, Browne-Wilkinson and Hoffmann agreed with the speeches of Lord Irvine of Lairg LC and Lord Steyn that the magistrates were entitled to determine the vires of the byelaw.)
98. There are a number of threads in the Law Lords’ reasoning. First and foremost there is the criminal context. The Law Lords objected strongly to the prospect of a defendant in a criminal case facing conviction for an offence and not being able to argue the vires of the law creating the offence. That, of course, can have no application in appeals before the Tribunal. Secondly, the Law Lords emphasised the practical difficulties facing a defendant of having to challenge the byelaw in separate judicial review proceedings and the potential duplication of proceedings. That may be a real difficulty in some cases but again, we would anticipate, would not necessarily be an insuperable obstacle for an individual who seeks to appeal to the Tribunal. Indeed, it is within our collective knowledge that concurrent judicial review proceedings are sometimes launched whilst an appeal is pending before the Tribunal (see, e.g. the ‘highly skilled migrant’ litigation: R(HSMP Forum Ltd) v SSHD [2008] EWHC 664 (Admin) and AA and others [2008] UKAIT 00003).
99. One final point. The Law Lords were concerned to dispel reliance by the prosecution upon the (then) championed ‘exclusivity rule’ in public law cases following O’Reilly v Mackman [1983] 2 AC 237. That rule, now fallen out of favour, was even then of no application where the challenge to the public law decision was raised as a defence (whether in a civil or criminal context) as it is difficult to sustain the argument that a defendant has abused the court’s process (the jurisprudential basis for O’Reilly) where he has not initiated the procedure (see Wandsworth LBC v Winder [1985] AC 461 and the cases referred to in Lord Steyn’s speech above). Of course, proceedings before the Tribunal are brought by the claimant and so the abuse of process argument might have some potential application here: but it does not. An appeal to the Tribunal is a statutory procedure specifically set up to allow challenge to an immigration decision. We do not see how in these circumstances initiating that procedure where part of the claim is that a statutory instrument is ultra vires can be properly described as an abuse of process (see also Foster, below per Lord Bridge of Harwich at p.762E-F).
100. We accept that Boddington demonstrates that there are situations where the vires of a statutory instrument (we see no distinction between that and the byelaw in Boddington) may be challenged outside of judicial review proceedings. However, beyond that, it does not seem to us that the decision materially advances Mr Draycott’s submissions on behalf of the Appellant. In our view, in Boddington the criminal context and practical difficulties facing a defendant of bringing judicial review proceedings were compelling and led the House of Lords to its conclusion. We see no basis for Mr Draycott’s submission that the claimant in this appeal is “in the nature of a defendant” in respect of the application of Art 33(2) of the Refugee Convention. The underlying rationale for the decision in Boddington does not drive us to reach a similar conclusion in respect of appeals before the AIT.
101. That said, we turn to the decision of the House of Lords in Chief Adjudication Officer v Foster which seems to present a more solid base for Mr Draycott’s submissions.
102. In Foster, the claimant, who was disabled, received certain social security payments including income support. The case concerned the claimant’s entitlement to ‘severe disability premium’ as part of her income support. That entitlement turned upon whether a part of the Income Support (General) Regulations 1987 (which excluded her entitlement) was ultra vires the enabling Social Security Act 1975. The claim was initially rejected by an adjudication officer and, on appeal, by the Social Security Appeal Tribunal. On further appeal, a Social Security Commissioner upheld her claim holding the relevant part of the 1987 Regulations to be ultra vires. On appeal from his decision, the Court of Appeal held that a Commissioner had no jurisdiction to consider the vires of the Regulation; that could only be decided in judicial review proceedings. The House of Lords overturned the Court of Appeal’s decision.
103. The House of Lords was concerned with the jurisdiction of the Commissioner on an appeal under s.101(1) of the 1975 Act which provided:
“An appeal lies to a Commissioner from any decision of a social security appeal tribunal on the grounds that the decision of the tribunal was erroneous in point of law.”
104. Lord Bridge of Harwich set out the context as follows (at p.762):
“the central question… is whether a claimant otherwise entitled to some social security benefit which has been denied to him by the adjudication officer and the appeal tribunal in reliance on some provision in a regulation which the Secretary of State had no power to make is entitled to succeed on appeal to the commissioner on the ground that the decision against him was 'erroneous in point of law' or whether, as must follow if the Court of Appeal were right, before he invokes the statutory machinery by which alone his claim can be enforced, he must first proceed by way of an application for judicial review to have the offending provision quashed or declared invalid. It is common ground that the principle of O'Reilly v Mackman [1982] 3 All ER 1124, [1983] 2 AC 237 has no application, since there can be no abuse of process by a party who seeks a remedy by the very process which statute requires him to pursue. It was further rightly accepted by [counsel for the Chief Adjudication Officer and secretary of State] before your Lordships that a decision giving effect to secondary legislation which is ultra vires is, indeed, in the ordinary meaning of the words 'erroneous in point of law'. The question then is whether, when that phrase is used in s 101 of the 1975 Act, there is something in the context in which it appears which requires by necessary implication that it be given a restricted meaning so as to exclude from its ambit any errors of law referable to a misuse by the Secretary of State of his regulation making power.”
105. Lord Bridge of Harwich (with whom Lords Templeman, Ackner, Browne-Wilkinson and Slynn of Hadley agreed) expressed the conclusion of the House of Lords on the vires point as follows (at pp.766-7):
“My conclusion is that the commissioners have undoubted jurisdiction to determine any challenge to the vires of a provision in regulations made by the Secretary of State as being beyond the scope of the enabling power whenever it is necessary to do so in determining whether a decision under appeal was erroneous in point of law. I am pleased to reach that conclusion for two reasons. First, it avoids a cumbrous duplicity of proceedings which could only add to the already overburdened list of applications for judicial review awaiting determination by the Divisional Court. Secondly, it is, in my view, highly desirable that when the Court of Appeal, or indeed your Lordships' House, is called upon to determine an issue of the kind in question it should have the benefit of the views upon it of one or more of the commissioners, who have great expertise in this somewhat esoteric area of the law.”
106. Mr Draycott placed considerable reliance upon Foster. He submitted that there is no discernible distinction between the Commissioners’ statutory jurisdiction that a decision is “erroneous in point of law” and the Tribunal’s jurisdiction to decide whether an immigration decision is “not in accordance with the law”. Mr Patel sought to distinguish Foster on a number of bases both in his oral submissions and his written submissions (see para 51 of written submissions dated 8 February 2008).
107. There are strong parallels between the statutory jurisdiction at issue in Foster and that of the AIT. It is difficult, if not impossible, to grasp any rational distinction between the statutory basis of an appeal phrased as “erroneous in point of law” (in the social security context) and one phrased as “not in accordance with the law” (in the case of the AIT). Mr Patel submitted that the latter presupposes that the “law” is that laid down in the statutory instrument. As we stated earlier, had it not been for Foster, we would have reached just this conclusion ourselves. However, in Foster the Law Lords took a diametrically apposed view as to the content of the word “law” in the social security legislation. It matters not, in our view, that counsel for the government in Foster conceded that “a decision giving effect to secondary legislation which is ultra vires is…in the ordinary meaning of the words ‘erroneous in point of law’” (see Lord Bridge of Harwich’s speech at p.762F). The House expressed no doubt about the correctness of the concession and subsequently Foster was followed on this point by the Court of Appeal in Howker v Secretary of State for Work and Pensions [2002] EWCA Civ 1263 where it was stated that the Commissioners’ jurisdiction was “not now in doubt” as a result of Foster (per Peter Gibson LJ at [33]).
108. Both Boddington and Foster require the statutory context to be considered in order to determine whether “by necessary implication” the 2002 Act was not intended to include “errors of law” generated by reliance upon an ultra vires statutory instrument. Mr Patel offered no reasoned basis upon which we could base a conclusion that the interpretation of “law” in Foster should be any different in an appeal under the 2002 Act.
109. In Foster, Lord Bridge of Harwich saw no objection to the Commissioners rather than the High Court reaching a decision on the vires of a statutory instrument. It was better that a “duplicity of proceedings” was avoided (pp.766H-787A). Likewise it was not problematic that, unlike the High Court, a decision of the Commissioners was only binding between the parties and had no wider force in law (at p.764G). Whilst we agree with the first of these points, we would consider the latter point to be an important argument supporting the Appellant’s submissions in this appeal. The effect of the 2004 Order is, in our view, better determined by the High Court whose decision will have general effect rather than being restricted to determining the outcome of a particular appeal before the AIT. However, the point is no better in the context of the AIT’s jurisdiction than in the social security context and so its rejection in Foster makes this an untenable basis for reaching a conclusion contrary to that in Foster.
110. It might be suggested that Lord Bridge of Harwich was content to reach the decision he did because, for him, the ultra vires issue was one which could be determined by the Commissioners alone – a small group of senior tribunal judges whom he described as “of comparable standing to circuit judges” (at p.762C). There would be, in other words, no prospect of a large number of first instance tribunals adjudicating upon difficult and, potentially at least, controversial and far-reaching legal issues relating to the vires of statutory instruments. If that were an important limitation, that might suggest a contextual difference with the AIT where the jurisdiction could not be so limited. It would be within the scope of any AIT Tribunal however constituted, not just those comprised of its senior judiciary, to determine the vires of a statutory instrument on the basis that an immigration decision was “otherwise not in accordance with the law”. However, whilst alive to the dangers, Lord Bridge of Harwich acknowledged in Foster that the jurisdiction to determine vires applied throughout the tribunal appeals hierarchy. He said (at p.762H):
“if the commissioner can base his decision in any case on the invalidity of some provision in regulations made under the Act, it must follow that appeal tribunals and adjudication officers can do likewise.”
111. Lord Bridge’s solution was to recognise the reality that any “serious challenge” to a statutory instrument was likely to find its way to the Commissioners on appeal (at p.763B-F). Precisely the same reasoning and outcome applies in the context of the AIT. Any such case is likely ultimately to come before a panel of senior AIT judges either specifically constituted at the initial hearing or on reconsideration when an immigration judge’s decision is challenged.
112. Consequently, it is difficult to avoid the conclusion as a result of Foster that if a decision-maker (or lower tribunal) in the social security context errs in law by applying “law” derived from an ultra vires statutory instrument, so too, it would seem, the decision-maker acts “not in accordance with the law” in applying ultra vires “law” in the immigration or asylum context. We recognise the significance of this if correct. It would not, however, be our view unless we were driven to reach it by Foster. For the reasons we are about to develop, it is not necessary for us to reach any concluded view in this appeal on the impact of Foster to the AIT’s jurisdiction because we have concluded that the 2004 Order is not in fact ultra vires the enabling power in s.72 of the 2002 Act.
2. Incorporation
113. Mr Draycott submitted that the Refugee Convention is itself part of English law. As such, the power in s.72(4) had to be exercised consistently with that Convention or any resulting statutory instrument would be ultra vires. He relied, in particular, upon s.2 of the Asylum and Immigration Appeals Act 1993 and the speeches of Lord Keith of Kinkel in R v SSHD ex parte Sivakumaran [1988] AC 958 and Lord Steyn in R(European Roma Rights Centre and others) v Immigration Officer at Prague Airport and another (UNHCR intervening) [2004] UKHL 55 for the proposition that the Refugee Convention had been incorporated into our domestic law. We do not agree that this is their effect.
114. In Sivakumaran, Lord Keith of Kinkel commented (at p.990) that:
“The United Kingdom having acceded to the Convention and Protocol, their provisions have for all practical purposes been incorporated into United Kingdom law.”
115. In the European Roma Rights Centre case, Lord Steyn, having cited Lord Keith of Kinkel’s statement in Sivakumaran, also referred to s. 2 of the Asylum and Immigration Appeals Act 1993 which is in the following terms:
“Nothing in the immigration rules (within the meaning of the 1971 Act) shall lay down any practice which would be contrary to the [Refugee] Convention.”
116. Lord Steyn continued (at [41]-[42]):
“41….It is necessarily implicit in section 2 that no administrative practice or procedure may be adopted which would be contrary to the Convention. After all, it would be bizarre to provide that formal immigration rules must be consistent with the Convention but that informally adopted practices need not be consistent with the Convention. The reach of section 2 of the 1993 Act is therefore comprehensive.
42. Parliament must be taken to have been aware, in enacting the 1993 Act, that the courts had treated references in the immigration rules to the Refugee Convention as "indirectly" or "for practical purposes" incorporating it into domestic law: Bennion, Statutory Interpretation, 4th ed (2002), p 469. In the context of the decisions of the Court of Appeal and House of Lords in 1987 Parliament must have intended that the strengthened reference to the Refugee Convention in primary legislation would be treated by the courts as an incorporation of the Refugee Convention into domestic law. Moreover, the heading of section 2 is "Primacy of the Convention." This is a relevant and significant pointer to the overriding effect of the Convention in English law: R v Montila and Others [2004] UKHL 50, paras 31-37, per Lord Hope of Craighead. It is true, of course, that a convention may be incorporated more formally by scheduling it to an enactment, eg the Carriage of Goods by Sea Act 1971 which enacted the Hague-Visby Rules. But there is no rule specifying the precise legislative method of incorporation. It is also possible to incorporate a treaty in part, e.g. the European Convention on Human Rights was incorporated into our law without article 13: see Human Rights Act 1998. In my view it is clear that the Refugee Convention has been incorporated into our domestic law.”
117. Lord Steyn’s final conclusion is strongly relied upon by Mr Draycott. Both Lord Bingham of Cornhill (at [7]) and Lord Hope of Craighead (at [50]) also referred to Lord Keith of Kinkel’s view on incorporation in Sivakumaran with apparent approval.
118. Despite the supportive nature of these dicta, we have considerable difficulty with the argument that anything in s.2 of the 1993 Act or elsewhere in our domestic legislation establishes that the Refugee Convention has become formally incorporated into English Law. It is trite law to state that a treaty must be expressly incorporated before it becomes part of English law. We were not shown any legislative provision purporting to do so beyond s.2 of the 1993 Act. On its face, s.2 has no application to this case which is concerned, not with the Immigration Rules, but with s.72 and the 2004 Order. Beyond that, s.2, which only applies to the Immigration Rules, would be unnecessary if the Refugee Convention was already part of our law.
119. We do not need to struggle further with the view of Lord Keith of Kinkel in Sivakumaran (and the apparent support given by Lord Steyn and the others in the European Roma Rights Centre case) because subsequently the legal position was put beyond doubt by the House of Lords itself in the case of R v Asfaw [2008] UKHL 31.
120. Asfaw concerned the interpretation of s.31 of the Immigration and Asylum Act 1999 which creates a statutory defence to certain listed immigration offences where the individual is a refugee. The defendant, who was a refugee in transit, was convicted of an offence relating to the presentation of a false passport but the offence was not one which was listed in s.31. The defendant sought to rely directly on Art 31 of the Refugee Convention which stated that penalties should not be imposed upon refugees in respect of their illegal presence or entry. One of the arguments presented to the House of Lords, but rejected by them, was that Art 31 was part of English law and so could be directly relied upon. In emphatically rejecting this submission Lord Bingham of Cornhill said (at [29]):
“29. The appellant sought to address this disparity by submitting that the Convention had been incorporated into our domestic law. Reliance was placed on observations of Lord Keith of Kinkel in R v Secretary of State for the Home Department, Ex p Sivakumaran [1988] AC 958, 990G; Lord Steyn in R (European Roma Rights Centre and others) v Immigration Officer at Prague Airport and another (United Nations High Commissioner for Refugees Intervening) [2004] UKHL 55, [2005] 2 AC 1, paras 40-42; section 2 of the Asylum and Immigration Appeals Act 1993; and para 328 of Statement of Changes in Immigration Rules (HC 395). It is plain from these authorities that the British regime for handling applications for asylum has been closely assimilated to the Convention model. But it is also plain (as I think) that the Convention as a whole has never been formally incorporated or given effect in domestic law. While, therefore, one would expect any government intending to legislate inconsistently with an obligation binding on the UK to make its intention very clear, there can on well known authority be no ground in domestic law for failing to give effect to an enactment in terms unambiguously inconsistent with such an obligation.”
121. At para [69], Lord Hope of Craighead, agreeing with Lord Bingham of Cornhill, added:
“The giving effect in domestic law to international obligations is primarily a matter for the legislature. It is for Parliament to determine the extent to which those obligations are to be incorporated domestically. That determination having been made, it is the duty of the courts to give effect to it.”
122. It is significant that both Lord Bingham of Cornhill and Lord Hope of Craighead who, in the European Roma Rights case had cited Lord Keith of Kinkel’s incorporation view in Sivakumaran, unambiguously rejected it in Asfaw. It is worth noting that in Asfaw the House was not concerned with the effect of the Qualification Directive (which we discuss below) presumably because it contains no equivalent of Art 31 of the Refugee Convention that could have been relied upon.
123. In the light of Asfaw, which was decided subsequent to the initial hearing before us in this appeal, Mr Draycott did not wholly abandon reliance upon his submission that the Refugee Convention had been incorporated into English law. However, he accepted that it had scaled down the broad proposition of Lord Steyn in the European Roma Rights Centre case.
124. In our view, Mr Draycott’s submission (scaled down or otherwise) is doomed to fail in the light of the clear and unchallenged statements of the law in the speeches of Lords Bingham of Cornhill and Hope of Craigheadin Asfaw. There is no suggestion that any of the other members of the House disagreed. As a result, in our judgment, the ‘incorporation’ of the Refugee Convention into English Law effected by s.2 of the 1993 Act is limited to that provision’s impact upon the content of immigration rules or any wider policy. Of course, the Tribunal must, when called upon to do so, deal with the argument that an individual’s removal in consequence of a particular immigration decision will be a breach of the Refugee Convention. That much follows from the statutory ground of appeal in s.84(1)((g) of the 2002 Act. Beyond that however, at its highest, reliance upon the Refugee Convention is confined to the established interpretative axiom that when construing legislation giving effect to a treaty obligation Parliament should be taken to have intended to give effect to that treaty’s terms unless clear contrary words are used. The problem faced by Mr Draycott in this appeal is that Parliament both in s.72 and the 2004 Order has clearly done just that. Parliament has unambiguously presumed to lay down a meaning of some of the words in the Convention, despite any autonomous international meaning that those words might have. No ordinary principle of statutory interpretation in English Law could led to a different reading of s.72(4) read with s.72(6) – the presumption that certain crimes are per se “particularly serious” ones is irrefutable.
125. For this reason, there is no basis – apart from the EU law argument below – which could lead us to the conclude that s.72 and the 2004 Order should not be applied precisely as the Panel in this appeal did.
126. We turn now to consider the effect of the Qualification Directive, in particular Art 21.
Art 21 of the Qualification Directive
127. In his skeleton argument, Mr Draycott placed reliance upon the Qualification Directive. As the case progressed, it became clear – as a result of some encouragement from the Tribunal – that this aspect of the case acquired greater significance in Mr Draycott’s argument. Ultimately, we received detailed written submissions from both parties on the effect of Art 21.2 of the Qualification Directive. We set out Art 21 above but for convenience we do so again here. It provides as follows:

“Protection from refoulement

1. Member States shall respect the principle of non-refoulement in accordance with their international obligations.

2. Where not prohibited by the international obligations mentioned in paragraph 1, Member States may refoule a refugee, whether formally recognised or not, when:

(a) there are reasonable grounds for considering him or her as a danger to the security of the Member State in which he or she is present; or
(b) he or she, having been convicted by a final judgement of a particularly serious crime, constitutes a danger to the community of that Member State.

3. Member States may revoke, end or refuse to renew or to grant the residence permit of (or to) a refugee to whom paragraph 2 applies.”
128. The importance of this provision is as follows. Mr Draycott submitted that Art 21.2 incorporates the non-refoulement provision in Art 33(2) of the Refugee Convention and therefore the ‘autonomous’ meaning of Art 33(2). He submitted that it would be contrary to Art 63 of the EC Treaty for the Qualification Directive to legislate in a manner incompatible with the Refugee Convention. The former should, therefore, wherever possible be interpreted so as to give effect to the Refugee Convention. Mr Draycott submitted that Art 21.2 is directly effective under EU law and thus ‘trumps’ any domestic legislation that is contrary to it. Thus, s.72 of the 2002 Act and the 2004 Order must either be read consistently with the ‘autonomous’ meaning of “particularly serious crime” or, under EU law, they can have no effect. Any issue of the Tribunal’s jurisdiction to hold a statutory instrument to be ultra vires is thereby avoided and, given the intrusive nature of EU law, even s.72 may be vulnerable. The Tribunal must give effect to EU law.
129. Mr Patel’s submissions on this aspect of the appeal were again straightforward. In his written submissions Mr Patel did not contend that Art 21 did not have direct effect. He accepted that the wording of Art 21 was, for all relevant purposes, identical to Art 33(2) of the Refugee Convention. However, he maintained his primary submission in relation to Art 33(2) – now part of EU law by virtue of Art 21 of the Qualification Directive – that s.72 and the 2004 Order were not contrary to any ‘autonomous’ meaning of Art 33(2) which he submitted did not exist.
130. We have already rejected Mr Patel’s primary submission on the proper meaning and interpretation of Art 33(2). That now places Mr Patel in some difficulties as to the effect of Art 21.2. We are wholly persuaded by Mr Draycott’s submissions on this aspect of the appeal. We have no doubt that Art 21.2 was intended to enshrine the non-refoulement obligation in Art 33(2) into the EU asylum regime. We did not understand this to be a matter of dispute between the parties. Art 21.2(b) is worded in identical terms to Art 33(2). The context makes plain that the Directive is seeking to give effect to the non-refoulement provisions of the Refugee Convention. Recitals (2) and (3) emphasis the central role played by the Refugee Convention in the EU’s protection regime. Recital (3) puts the Refugee Convention at the heart of the EU asylum system, including the Qualification Directive:

“The Geneva Convention and Protocol provide the cornerstone of the international legal regime for the protection of refugees.”
131. Recital (2) deals specifically with refoulement:

“The European Council at its special meeting in Tampere on 15 and 16 October 1999 agreed to work towards establishing a Common European Asylum system, based on the full and inclusive application of the Geneva Convention relating to the Status of Refugees of 28 July 1951 (Geneva Convention), as supplemented by the New York Protocol of 31 January 1967 (Protocol), thus affirming the principle of non-refoulement and ensuring that nobody is sent back to persecution.”
132. As to Art 21.2’s direct effect, Mr Patel accepted in his written submissions that it has direct effect. He was right to do so. That is the view we would have taken had it been a matter of dispute between the parties given that Art 21.2 is “unconditional” and “sufficiently precise” (see, e.g. Marks & Spencer plc v Customs and Excise Commissioners (Case C-62/00) [2003] 2 WLR 665 at [24]-[26]). Once that is accepted, it follows that s.72(4) and the 2004 Order must either be read consistently with the ‘autonomous’ meaning of Art 33(2) or simply not applied (see, e.g. Marleasing SA v La Comercial Internacional de Alimentacion SA (Case C-106/89) {1990] ECR I-4135).
133. This approach governs whether the domestic legislation is seeking to give effect to the Directive (which s.72 and the 2004 Order are not) or, as in this appeal, where directly enforceable EU rights arise independently. The latter is exemplified by the decision of the House of Lords’ in Autologic Holdings plc v IRC [2005] UKHL 54 to which we were referred by Mr Draycott. That case concerned the compatibility of provisions in the Income and Corporation Taxes Act 1988 with fiscal reliefs and advantages derived from Community law rights. The principal issue was whether the Special Commissioners should give primacy to those EU rights. In holding that they should, Lord Nicholls of Birkenhead set out the approach to be adopted (at [16]-[17]):
“16. The second basic principle concerns the interpretation and application of a provision of United Kingdom legislation which is inconsistent with a directly applicable provision of Community law. Where such an inconsistency exists the statutory provision is to be read and take effect as though the statute had enacted that the offending provision was to be without prejudice to the directly enforceable Community rights of persons having the benefit of such rights. That is the effect of section 2 of the European Communities Act 1972, as explained by your Lordships' House in R v Secretary of State for Transport, Ex p Factortame Ltd [1990] 2 AC 85, 140, and Imperial Chemical Industries Plc v Colmer (Inspector of Taxes) (No 2) [1999] 1 WLR 2035, 2041.
17. Thus, when deciding an appeal from a refusal by an inspector to allow group relief the appeal commissioners are obliged to give effect to all directly enforceable Community rights notwithstanding the terms of sections 402(3A) and (3B) and 413(5) of ICTA. In this regard the commissioners' position is analogous to that of the Pretore di Susa in Amministrazione delle Finanze dello Stato v Simmenthal SpA (Case 106/77) [1978] ECR 629. Accordingly, if an inconsistency with directly enforceable Community law exists, formal statutory requirements must where necessary be disapplied or moulded to the extent needed to enable those requirements to be applied in a manner consistent with Community law. Paragraph 70 of Schedule 18 to the Finance Act 1998 is an instance of such a requirement. Paragraph 70 provides that a claim for group relief requires the consent of the surrendering company, which must be given by notice in writing to its own inspector of taxes when or before the claim is made. This provision cannot be applied literally in the case, say, of a German subsidiary which makes no tax returns in this country. So if the residence restriction is found to be inconsistent with Community law this provision will need adapting so as to give effect to the overriding Community rights. In this regard the appeal commissioners have the same powers and duties as the High Court.” (emphasis added)
134. We considered earlier the proper construction of s.72 of the 2002 Act. Applying accepted canons of construction applicable to English legislation, it is impossible to read s.72(4) as creating anything other than an irrebuttable presumption that any offence specified in the 2004 Order is a “particularly serious crime”. The same would follow for the presumption in s.72(2) where the individual is convicted of an offence for which he receives a sentence of imprisonment of at least 2 years and also in s.72(3) in respect of the equivalent conviction overseas. Section 72(6) provides:
“A presumption under subsection (2), (3) or (4) that a person constitutes a danger to the community is rebuttable by that person.”
135. The legislative intent is manifest: only the presumption that the individual is a “danger to the community” can be rebutted. In our judgment, the preliminary view of the Court of Appeal in TB (Jamaica) to this effect is correct. However, the force of the interpretative canon of EU law requires us to construe and apply s.72(6) as if it included the presumption that a conviction for an offence listed in the 2004 Order is one for a “particularly serious crime”. Section 72(6) must be radically “moulded” (to use Lord Nicholls’ word) in this way otherwise, short of disapplying s.72 (and the 2004 Order), we see no other way of giving effect to the EU obligation founded in Art 21.2 of the Qualification Directive.
136. It follows that the Panel in this appeal erred in law in paragraph 17 of its determination (rejecting the Appellant’s submissions to the contrary) when it held that the Appellant could not as a matter of law rebut the presumption that his conviction of the offence under s.3(1) of the Sexual Offences Act 2003 was a “particularly serious crime” by virtue of the combined effect of s.72(4) of the 2002 Act and the 2004 Order.
137. That said we do not consider that this error affected the outcome of the appeal. Despite prompting from us, Mr Draycott did not suggest any basis upon which the Appellant’s offence would not properly be characterised as a “particularly serious crime” other than to submit that it was “self-evidently” not sufficiently serious to meet the high threshold of Art 33(2). We disagree. We are in no doubt that the offence of sexual assault contrary to s.3(1) of the 2003 Sexual Offences Act is capable of being a “particularly serious crime”. The nature of the offence and the maximum penalty of 10 years imprisonment on indictment are indicative of that. The actual sentence passed of 21 months imprisonment and a recommendation for deportation also points in that direction. So far as we can tell, the offence involved the touching of a 20 year old woman in circumstances where the Appellant maintained his innocence and blamed the victim for the offence having been committed after they consumed alcohol together. So much is clear from a Probation Officer’s report which is in the appeal file which also notes that the Appellant’s “remorse is low”. The jury clearly disagreed with the Appellant’s myopic view of his own responsibility and concluded that the victim had not consented to the sexual touching and the Appellant did not reasonably believe she was consenting (both requirements of the offence) when they returned a guilty verdict.
138. As we have said, Mr Draycott did not put forward any argument to rebut the presumption that on these facts, the Appellant had not been convicted of a “particularly serious crime”. The Panel’s finding that the Appellant had not on the evidence rebutted the presumption that he was a “danger to the community” was not challenged before us. We can see no basis upon which the decision of this Tribunal can be other than to agree with the Secretary of State’s certificate under s.72(9) that the presumptions in subsection (4) applied and thus dismiss the appeal on asylum grounds.
139. For the same reasons, the Panel’s decision that the Appellant is excluded from humanitarian protection by virtue of para 339D(i) of the Immigration Rules is unassailable in law. There are “serious reasons” for considering that the Appellant has committed a “serious crime”. On the basis of the Panel’s findings, the same would follow applying para 339D(iii), namely that the Appellant constitutes a “danger to the community”.
Credibility

140. In his written and oral submissions, Mr Draycott mounted a sustained attack upon the Panel’s rejection of the Appellant’s credibility and of the whole of his account of what the Appellant claimed happened to him during military service in Eritrea and which led him to escape to Sudan in May 2000. As a consequence, he submitted that the Panel materially erred in law in dismissing the Appellant’s appeal under Art 3 of the ECHR.

141. Mr Draycott made a number of general and specific criticisms of the determination. First, the Panel unduly relied upon its view about the plausibility of various aspects of the Appellant’s evidence in rejecting it – in the words of Mr Draycott, the Panel rested its views on “Western standards” rather than those pertaining to people in Eritrea. Secondly, the Panel failed adequately or at all to assess the Appellant’s evidence in the light of the objective material that was placed before it. Thirdly, Mr Draycott criticised the Panel for failing to take account of the medical evidence that was before it relating to injuries which the Appellant claimed were caused whilst engaged in military service in Eritrea. Mr Draycott submitted that the Panel effectively reached its conclusion on credibility before it looked at that medical evidence in paragraph 23 of its determination. Finally, Mr Draycott submitted, relying upon specific instances of what he had identified as general errors of approach by the Panel, that the Panel had erred in law in its approach to the Appellant’s evidence in respect of specific events which the Appellant claimed had occurred whilst serving in the Eritrean military.
142. Mr Patel more or less accepted that there were some errors that could be identified in the Panel’s determination. However, in his submission, much of the Panel’s reasoning in rejecting important aspects of the Appellant’s evidence was unimpeachable in law and justified the Panel’s overall conclusion to find the Appellant not to be credible and to reject his account.
143. We deal first with Mr Draycott’s general criticisms of the Panel’s approach before turning to the detailed challenge he made to specific parts of their reasoning.
144. Mr Draycott’s first point concerns the Panel’s recourse to the notion of “implausibility” of the Appellant’s account to reject it. We were referred to the observations of Sedley LJ in A v SSHD [2006] EWCA Civ 973 at [7] as follows:
“7. I have to say that the now prevalent practice of finding that an applicant is lying because the events described by him are “implausible” is not attractive and may even be said not to be intellectually respectable. We all know from our own lives that the improbable and the implausible happed repeatedly. It is the task of the fact-finder to decide not whether a particular occurrence or set of occurrences is probable or plausible in the sense that it was likely to happen, but whether, however objectively improbable or implausible, it is what did happen. Increasingly, as it seems to me, immigration judges are substituting the former for the latter.”
145. These remarks were made by Sedley LJ in the context of granting permission to appeal to the Court of Appeal. Even assuming that these are properly citable before us, they cannot, nor in our view were they intended to, be anything more than entering a cautionary note about the approach of decision-makers and their reliance upon the implausibility of an account as a basis for rejecting it. We would reiterate that caution. On the other hand, it cannot be said as a matter of principle that a decision-maker cannot properly rely upon the implausibility of an account or, the converse, its plausibility in assessing the veracity of a witness or the cogency of the evidence. For example, if an individual in his evidence was to claim that he travelled from London to Edinburgh by foot in an afternoon the implausibility (indeed impossibility) that what he is saying could occur is something that the decision-maker may properly take into account, in this case, to reach the conclusion that his evidence should not be accepted. The important point, particularly in respect of appeals heard by the AIT, is that the assessment of an individual’s evidence (often the Appellant’s) will relate to what is claimed to have happened in another country where what may seem implausible here may be altogether less implausible in the circumstances pertaining in that other country.
146. The difficulties of fact-finding in asylum cases and the proper approach to be taken is, in our view, identified by the Court of Appeal in the case of HK v SSHD [2006] EWCA Civ 1037 at [27]–[30]. There, Neuberger LJ said this:
“27. The difficulty of the fact-finding exercise is particularly acute in asylum cases, as has been said on more than one occasion in this court - see for instance Gheisari –v- Secretary of State [2004] EWCA Civ 1854 at paragraphs 10 and 12 per Sedley LJ and at paragraphs 20 and 21 per Pill LJ. The standard of proof to be applied for the purpose of assessing the appellant's fear of persecution is low. The choice is not normally which of two parties to believe, but whether or not to believe the appellant. Relatively unusually for an English Judge, an Immigration Judge has an almost inquisitorial function, although he has none of the evidence-gathering or other investigatory powers of an inquisitorial Judge. That is a particularly acute problem in cases where the evidence is pretty unsatisfactory in extent, quality and presentation, which is particularly true of asylum cases. That is normally through nobody's fault: it is the nature of the beast.
28. Further, in many asylum cases, some, even most, of the appellant's story may seem inherently unlikely but that does not mean that it is untrue. The ingredients of the story, and the story as a whole, have to be considered against the available country evidence and reliable expert evidence, and other familiar factors, such as consistency with what the appellant has said before, and with other factual evidence (where there is any).
29. Inherent probability, which may be helpful in many domestic cases, can be a dangerous, even a wholly inappropriate, factor to rely on in some asylum cases. Much of the evidence will be referable to societies with customs and circumstances which are very different from those of which the members of the fact-finding tribunal have any (even second-hand) experience. Indeed, it is likely that the country which an asylum-seeker has left will be suffering from the sort of problems and dislocations with which the overwhelming majority of residents of this country will be wholly unfamiliar. The point is well made in Hathaway on Law of Refugee Status (1991) at page 81:
"In assessing the general human rights information, decision-makers must constantly be on guard to avoid implicitly recharacterizing the nature of the risk based on their own perceptions of reasonability."
30. Inherent improbability in the context of asylum cases was discussed at some length by Lord Brodie in Awala –v- Secretary of State [2005] CSOH 73. At paragraph 22, he pointed out that it was "not proper to reject an applicant's account merely on the basis that it is not credible or not plausible. To say that an applicant's account is not credible is to state a conclusion" (emphasis added). At paragraph 24, he said that rejection of a story on grounds of implausibility must be done "on reasonably drawn inferences and not simply on conjecture or speculation". He went on to emphasise, as did Pill LJ in Ghaisari, the entitlement of the fact-finder to rely "on his common sense and his ability, as a practical and informed person, to identify what is or is not plausible". However, he accepted that "there will be cases where actions which may appear implausible if judged by…Scottish standards, might be plausible when considered within the context of the applicant's social and cultural background".”
147. In part therefore, Mr Draycott’s first general criticism of the Panel’s determination merges with his second, namely that the Panel failed to assess the Appellant’s evidence in the light of the objective evidence before it.
148. It is true that the Panel refers to the Appellant’s evidence as being “implausible”, “intrinsically implausible”, “extraordinary” or “surprising” over 20 times in a somewhat dense and often unstructured effusion of lengthy paragraphs; all of which makes this determination difficult to read, let alone analyse. However, in our view, the proper criticism, if any, of the Panel’s recourse to “implausibility” must lie in its approach to particular aspects of the Appellant’s evidence for the reasons we have already given.
149. Turning to Mr Draycott’s second criticism, the Panel had before it a number of bundles prepared by the Appellant’s representatives including a bundle containing a section headed “Objective Evidence” beginning at page 202 of that bundle and ending at page 304. We see nothing in Mr Draycott’s general criticism of the approach of the Panel to the objective evidence. It is clear to us that the Panel had well in mind the need to take account of the objective evidence. At paragraph 23 of its determination, the Panel specifically records that it has “taken into account the background material and objective evidence which has been submitted”. Again, the general criticism can only have any effect on the sustainability of this determination as a matter of law by looking at the way the Panel dealt with the specific aspects of the Appellant’s evidence and whether it can be said they adequately had regard to the background evidence that might have been relevant.
150. Finally, in relation to Mr Draycott’s general criticisms, we see nothing in his submission that the Panel failed to have regard to the medical evidence before it relating to the Appellant’s injuries in assessing the Appellant’s credibility. In paragraph 23 of its determination the Panel said this:
“23. We have considered the statements by the general practitioner and consultant which have been submitted. We note the possible explanation attributed by the general practitioner referred to at page 111. We note the analysis where injuries could be consistent with descriptions of injury given by the Appellant. We note that the general practitioner Dr Phythian-Adams could not comment on the mechanism of fracture in relation to the fractured bone in the Appellant’s left hand which the Appellant described. We note that the consultant plastic and reconstructive surgeon, Mr Mark Ho-Asjoe stated that it was unfortunately difficult to suggest any form of aetiology to the scars referred to at page 119. There was multiple scarring relating to his profession. It has been accepted by the respondent that the Appellant was in the military. From the scar point of view, the author stated it was difficult to associate any form of aetiology to the scarring because of the regular repetitive damage to the medial and inferior aspect of the feet from walking. We do not find that the medical evidence submitted bears out the causation claimed by the Appellant on the basis of his description of the events which have befallen him.”
151. Of course, medical evidence relevant to the veracity of the Appellant which, for example, confirms or, more likely, is consistent with events claimed to have happened must be taken into account as part of the overall assessment of the Appellant’s credibility. So much is clear from the decision of the Court of Appeal in Mibanga v SSHD [2005] EWCA Civ 367 and the many cases which have followed it. It is clear from paragraph 23 that the Panel did precisely this. Self-evidently, there the Panel took the medical evidence into account before reaching its final conclusion on the Appellant’s credibility in the final sentences of paragraph 23. Also, in our judgment, the Panel was entitled to give the weight that they did to that medical evidence, namely that it did not bear out the causation claimed by the Appellant. In context, the Panel obviously meant that, whilst it might be consistent with the Appellant’s account, the medical evidence (as patently was the case) did not establish, to the exclusion of other possibilities, the aetiology of the Appellant’s injuries. Taken into account as part of the overall evidence, there can be nothing objectionable in this. We see no discernable error of law in the Panel’s approach here.
152. Turning now to the detailed criticisms made of the Panel’s determination and its conclusion to reject the Appellant’s credibility, we focus first upon those aspects of the Appellant’s claim, which, if accepted, would put him in the ‘at risk’ category of a military deserter who has illegally left Eritrea. It was not disputed before us, indeed it could not be in the light of the Tribunal’s country guidance case MA (Draft evader – illegal departures – risk) Eritrea CG [2007] UKAIT 00059, that if the Appellant’s account was true he is at risk of serious ill-treatment contrary to Article 3 of the ECHR if he returned to Eritrea. (Of course, that would also justify findings that he would be at risk of persecution or of ill-treatment justifying humanitarian protection but those matters were not (properly in our view) before the Panel, because of the effect of s.72 of the 2002 Act and paragraph 339D(i) of the Immigration Rules).
153. As we have said, the form of this determination makes comprehension and analysis not without its difficulties. Mr Draycott’s detailed submission sought to attack a series of negative credibility findings in relation to specific events which formed part of the Appellant’s narrative. The crux of the Appellant’s account which, he said, put him at risk if returned to Eritrea is that he was a conscript in the Eritrean military having been called up (for a second time) in May 1998. There are then, in our view, three central events upon which he relies to establish his case. First, he says that a few months after he was called up he attended a military conference where he spoke out against and criticised the Eritrean authorities and the war with Ethiopia. As a result of this, three days later he was detained in prison, accused of being a spy, tortured before being released some six months later in February 1999 when he was sent to rejoin the fighting. The essentials of the Appellant’s evidence in respect of this are set out at paragraphs 11 – 20 of his statement dated 30 August 2006 (at pp.1-9 of the Appellant’s main bundle).
154. Secondly, the Appellant says that in April 2000, when he discovered that his wife and daughter were to be deported to Ethiopia, he questioned this and was imprisoned in Barentu prison, a high security prison as the Appellant describes it. He was in prison for eleven days before the Ethiopian troops overran Barentu and the prison guards fled. The Appellant then claims that he escaped from the prison opening his handcuffs with a nail and breaking the padlock on the door by hitting it with the handcuffs and fled to Sudan. The third central part of his claim is that he then fled to Sudan by walking the 120 -130 kilometres over a three day period.
155. The Panel rejected the Appellant’s account in respect of each of these three events.
156. As regards the first, Mr Draycott criticised aspects of the Panel’s reasoning in paragraphs 18 and 19 of its determination. In particular, he criticised the Panel for concluding that it was “extraordinary “that there should be a three day delay between the conference and the Appellant being detained. Mr Draycott submitted that the Panel had failed to have regard to an Amnesty International letter of 27 May 2002, which was considered and adopted, by the European Court of Human Rights in the case of Said v Netherlands (2006) 43 EHRR 248 (at tab 4 of the Appellant’s bundle of authorities before the Panel) where it is stated:
“It was usual for the Eritrean army to get together after an offensive and to conduct an evaluation of that offensive. It was also not unusual for a considerable time to pass between openly expressed criticism and arrest, or for deserters to be punished by their superiors without trial”.
157. Mr Draycott submitted that the Panel had failed to have regard to this in reaching their view that the Appellant’s evidence was “extraordinary”.
158. Also, Mr Draycott criticised the Panel where in paragraph 18 of its determination it said:
“Given the nature of the accusations against the Appellant and his own account of what he said at the big military conference we find it astounding that following his detention he was sent back to his old division”.
159. Mr Draycott referred us to paragraph 5.47 of the COI Report of 28 April 2006 (which is at tab 18 of the Appellant’s main bundle which was before the Panel) where it is stated, referring to the US State Department Report for 2005:
“There were substantial reports that prison conditions for persons temporarily held for evading military service were also poor. Unconfirmed reports suggested there maybe hundreds of such detainees. Draft evaders were typically held for one to twelve weeks before being reassigned to their units.”
160. Mr Draycott submitted that the Panel’s failure to take this evidence into account, given that it supported the Appellant’s account, undermined in law the Panel’s rejection of that evidence.
161. Further, Mr Draycott criticised the Panel’s view that it was “extraordinary” (at paragraph 19 of its determination) that following his detention the Appellant was returned to active service and did fighting in the front line given his claimed injuries which had occurred during his detention. Mr Draycott referred us to the case of KA (Draft- related risk categories and updated) Eritrea CG [2005] UKAIT 00156 at [81], the Human Rights Watch Report of 18 January 2006 and the COI Report for 28 April 2006 at para 6.19 which, he said, identified the often adverse “conditions of service” for conscripts. Mr Draycott reminded us that the Appellant’s evidence was that he had suffered a back injury during his detention and that there was no suggestion that he was unable to continue his military service because of it.
162. Mr Draycott also criticised the Panel’s view that it was “astonishing” that the Appellant would be returned to the front-line following his detention given his evidence that he had completed a form on joining up which showed his mother was Ethiopian. Mr Draycott pointed out that the Appellant was an Eritrean citizen (see COI Report, para 5.05) and there was no evidence that those with Ethiopian parents were exempt from military service. All Eritrean citizens aged 18 to 50 were eligible (see COI Report, para 5.58).
163. We have considerable misgivings about the Panel’s approach to the Appellant’s evidence leading up to and including his detention and release between May 1998 and February 1999. It is always a matter for the fact-finder, subject to an over-arching yardstick of perversity, to weigh up evidence and reach findings and conclusions based upon it. The fact-finder must, however, take account of all the relevant evidence in doing so. We accept Mr Draycott’s submissions that the Panel failed adequately, or at all, to do so in this case.
164. In reaching its view that the Appellant’s evidence was “extraordinary” and “implausible” in respect of his detention in May 1998 and his release and return to active duty at the front in February 1999, the Panel in our judgment failed to have regard to the background evidence supporting the substance of the Appellant’s account, in particular to which we have been referred. In addition, the Panel misunderstood, or at least mischaracterised, the nature and seriousness of the Appellant’s claimed injuries and his ability, therefore, to return to active duty. He did indeed claim his back was damaged and painful but the background evidence shows the harsh conditions imposed upon conscripts especially at times of active fighting between Eritrea and Ethiopia. Likewise, that the Appellant’s mother was Ethiopian cannot, in the light of the background evidence, justify the incredulity of the Panel that he would not be returned to the front-line. Indeed, neither can his claim that in February 1999 as a result in an escalation of hostilities with Ethiopia those who had been previously detained were released in order to continue their Eritrean military service.
165. Turning now to the Appellant’s second detention in and escape from Barentu prison in May 2000, the Panel’s reasons for rejecting this aspect of the Appellant’s account can be found at paragraph 20 of its determination:
“We find the account given by the Appellant of the circumstances of his escape from Barentu prison to be intrinsically implausible. In his witness statement dated 5th May 2004 the Appellant described the Eritrean military retreat after which the Appellant and the others broke the doors of the prison and they fled. It was a high security prison as the Appellant has described at paragraph 13. The Appellant repeats this description at paragraph 27 of his statement dated 30th August 2006. The people who are there were to be killed in secret. Again the Appellant refers to breaking the doors of the prison with the others and fleeing. However in the SEF interview in answer to question 56 the Appellant described the prison. He stated it was an underground prison and there was only one door. There were guards at the door. They are there for 24 hours in turn. The Appellant has described one of the inmates being handcuffed so they managed to unlock by a nail and then they unlocked the handcuffs and then broke the padlock on the door by hitting it with the handcuffs. We find this description to be intrinsically implausible. The Appellant amplified his explanation in the proceedings before us. We have set this out above. Despite the lack of technological barriers we find it astounding that a handcuff unlocked by means of a nail and which featured spikes would be capable of use to break down the single door of an underground prison classified as a high security institution. The guards had already left. Nonetheless the physical achievement described by the Appellant we find intrinsically implausible. We note the Appellant has stated that while he described the prison as a high security prison it was only considered such because there were many guards present to prevent escape. The actual technology was very poor. In that explanation to be found at pages 12 and 13 of the Appellant’s witness statement dated 30th August 2006 the Appellant refers to being given the opportunity because all the guards had fled at this time to break though the doors of the cell.”
166. Mr Draycott submitted that the Appellant’s evidence was supported by a CNN internet report dated 18 May 2000 which was produced to the Panel at the date of the hearing. We were not taken directly to this report but, as we understand it, the substance relied upon is set out in the grounds of review as follows:
“Ethiopian troops Wednesday captured the town of Barentu, about forty five miles north of the border.
Wrku Tesfamichael, Director of the state run Eritrean Relief and Refugee Commission, said Ethiopia’s military campaign already had driven a total of 550,000 Eritreans from their homes or the temporary camps where they were sheltered”.
167. Mr Draycott submitted that the Panel, in finding that the Appellant’s account of his imprisonment and escape from Barentu prison was “intrinsically implausible”, had failed to have regard to this supporting evidence. We do not accept this. It is clear that the Tribunal was well aware of this evidence. At paragraph 23 of its determination the Panel stated:
“We have noted in particular the evidence obtained by Counsel for the Appellant from the internet in relation to the fall of Barentu.”
168. We accept that the CNN report is not referred to in paragraph 20 of the Panel’s determination when considering the evidence of the Appellant in relation to his imprisonment in Barentu. However, there is no doubt in our view that the Tribunal had in mind the evidence about the fall of Barentu to the Ethiopians around the time that the Appellant claimed that he escaped when considering his evidence. Reference to the fall of Barentu was made in the submissions on behalf of the Appellant made to the Tribunal. These are fully recorded at paragraph 16 of its determination where the Tribunal said:
“We were referred to the 18th May report. We were referred to the US State Department Report 2001. This stated that Barentu fell to the Ethiopian troops. There was a coincidence to this being so close to the Appellant’s account.”
169. In our judgment, it would be wrong to conclude that the Panel was not aware of this evidence supporting the Appellant’s account of what happened in May 2000 and failed to take it into account in reaching its conclusion. We reject Mr Draycott’s submissions in this respect.
170. In addition, Mr Draycott criticised the Panel’s reasoning about the method of escape from Barentu prison employed by the Appellant. He submitted that the Panel erred in law in concluding that it was intrinsically implausible that the Appellant would open his handcuffs using a nail and then break open the padlock on the door to the prison by hitting it with those handcuffs. Mr Draycott submitted that the Panel’s conclusion was unsustainable. There was evidence before them of prison breaks one of which showed an escape by means which might otherwise seem as being implausible, namely where a wall had been pushed over and troops had been killed during the escape.
171. We do not accept Mr Draycott’s submissions. The Panel had regard to the situation in which the Appellant found himself, namely incarceration in a high security underground prison. In our view, the Panel was entitled to reach the view that the method of escape described by the Appellant was implausible, in the sense of unlikely and on the basis of that to reach the conclusion that his evidence in this respect was not to be accepted. We see no error of law in the approach or reasoning of the Panel here. The Panel was entitled to reach the view that it did.
172. The third aspect of the Appellant’s account crucial to his claim for international protection in the UK concerns his flight to Sudan once he escaped from prison. In paragraph 21 of its determination, the Panel rejected this aspect of the Appellant’s evidence. Its reasons were as follows:
“21. The Appellant has described taking around three days on foot having escaped Barentu prison to reach Kassala in Sudan. Prior to the Appellant’s transfer to Barentu prison the Appellant had been tied with a rope by the division administrator and taken to the prison. This punishment, the Appellant states, is called “helicopter.” Despite this treatment, the back injury which the Appellant had sustained during his period of about six months’ detention and the debilitating circumstances of his detention in Barentu prison which the Appellant has related he was nonetheless able to cover a very considerable distance to Sudan from Barentu on foot. During the hearing the Appellant submitted a note that the distance from Barentu to Kassala is about 120-130 kilometres and the climate was very hot so that it was impossible to walk for a long time at any one time. Despite the climatic conditions and the physical condition of the Appellant he was clearly, it appears, able to cover some 40 kilometres each day. We find this intrinsically implausible in relation to someone who had suffered the physical privations and injuries described by the Appellant. The Appellant stayed in the Sudan from May 2000. It was only in February 1999 that the Eritrean government, on the Appellant’s account, had been forced to release the ex-soldiers in prison and make them including him join the fighting. The Appellant was sent back to his old division. In his witness statement dated 30th August 2006 the Appellant refers being sent to the hospital in Asmara in March 2000 while he had problems with his back and caught malaria and was very ill. His wife and child visited him. In his witness statement dated 20th April 2004 the Appellant refers being given permission to stay with his fiancée for a week. He went back to the military camp after a period of one week as expected. Thereafter he was taken to another military camp and stayed at this military camp until April 2000. At the end of that month he was told by his sister his wife and daughter had been deported. There is no mention in that witness statement of the Appellant going to hospital in March 2000. In the Appellant’s supplementary witness statement of 5th May 2004 the Appellant refers at paragraph 11 to the marriage ceremony and being given permission to stay with his wife for a week. He was allowed to stay with her in another camp in the military front. In March 2000, he states, he received a letter from his sister saying his wife had been deported with his daughter. Straight away the Appellant went to the division administrator and told him there was no reason for him to continue fighting. After that the Appellant was tied up and taken to Barentu prison. There was no mention in this witness statement of the Appellant being in hospital in March 2000. We find that the Appellant has further embroidered his account. We find that the significance of the Appellant’s receiving treatment in March 2000 is to assist in explaining how the Appellant was able to reach the Sudan having escaped from Barentu prison. Without being in hospital in March 2000 the Appellant would have received no treatment for his back and would have endured the difficulties originating with his detention for about six months following his speech at the military conference and having been sent back to fight without any or proper attention. We do not find it plausible that the Appellant would have been able on the account he has given including of his treatment and injuries and wounds that he would have been able to accomplish such an arduous and lengthy journey in the time he has claimed to the Sudan. Plainly the Appellant has realised this difficulty in the time which has elapsed from his original statement and supplementary statement to which we refer of 2004 and has attempted to improve the level of plausibility of what is already an intrinsically implausible account. Had the appellant suffered and continued to suffer such back problems in March 2000 it is even more extraordinary that he was able to accomplish the journey to the Sudan.”
173. Mr Draycott criticised this reasoning. First, he submitted that the Panel had wrongly overstated the nature of any injuries suffered by the Appellant prior to his flight in reaching their conclusion that it was “intrinsically implausible” that someone with those injuries could make the journey from Barentu to Kassala in Sudan in the claimed three days. Secondly, and linked to that, Mr Draycott criticised the Panel for taking into account what it considered to be in effect an embellishment of his account that he had been hospitalised in March 2000.
174. There is no doubt that the Panel was well aware of the medical evidence relating to the Appellant’s injuries. As we noted earlier the evidence is referred to by the Panel at paragraph 23 of its determination. It does seem to us, however, that the Panel appears to regard the Appellant’s claimed back injury as inconsistent with his ability to “accomplish such an arduous and lengthy journey in the time he has claimed to Sudan.” We are not clear why the Panel took such a view. As we noted earlier, the Panel seems to have overstated or mischaracterised the Appellant’s claimed injuries.
175. Equally, the Panel’s criticism of the Appellant for not referring to his hospitalisation in March 2000 in his witness statements in April and May 2004 is somewhat diluted by the fact that he does refer to it in his SEF interview on 25 May 2004. It is not the case, therefore, as the Panel appeared to think that the Appellant first referred to this in his witness statement of 30 August 2006. At Question 41 of his interview when asked whether he received any medical treatment for injuries he had suffered whilst he was first detained, the Appellant replied:
“In March 2000, I went to a hospital because my back bone started to deteriorate and I was poorly with malaria.”
176. Nevertheless, it remains the case that the Appellant claimed to have walked about 120-130 kilometres in three days in a climate which in his own words “was very hot so that it was impossible to walk for a long time at anyone time”. It seems to us that given the evidence that the Appellant had suffered a back injury which, he claimed, subsequently showed in an x-ray taken in Sudan that his backbone was bent and, given the climatic circumstances acknowledged by the Appellant, the Panel was entitled to reach the view that it was implausible that the Appellant (trained soldier or not) could, in those circumstances, walk seventy-five miles in a three day period. In the result, therefore, we see no discernable error of law in the Panel’s rejection of the Appellant’s evidence concerning his flight to Sudan.
177. Mr Draycott also relied upon a number of other matters which he submitted had wrongly led the Panel to find the Appellant’s account incredible.
178. Mr Draycott referred us to paragraph 22 of the determination where the Panel considered the Appellant’s evidence about his stay in Sudan between May 2000 and April 2004 when he came to the UK. The Panel disbelieved him:
“22. The Appellant arrived in Sudan in May 2000 and only left on 6th April 2004. In his screening interview at page A12 the Appellant stated that he tried to claim asylum in Sudan but they would not accept an ex-soldier. In his witness statement dated 20th April 2004 the Appellant stated I did not apply for asylum in Sudan as I feared I will be deported back to Eritrea like many other people were being deported. In the SEF interview the Appellant explained that his friends had been asked by him about asylum and they told him not to claim. He also explained that if you are an ex-soldier they do not accept you as a refugee because of strained relationship with Eritrea and secondly you are an ex-soldier and if they send you into a refugee camp you do not need military training to join the Islamic Movement which is a big plus. Given the reasons that the Appellant has set forward for not claiming asylum in Sudan namely that he was told by his friends that if you claim asylum first they will take you into a refugee camp and then they will force you to join the Islamic Movement against Eritrea so he did not want to get into politics again and war we find it very surprising that the Appellant spent so long in Sudan. He was separated from his wife and child. He had been told to wait effectively until 2003 in relation to his desire to go to America. He did not leave the Sudan until 2004. Had the Appellant truly been at risk we find that he would not have spent so long in the Sudan having made up his mind that he could not claim asylum there. We find that the reality of the situation is that the Appellant was determined to go to America and that was his priority. We find that the Appellant is using the United Kingdom as a transit point for his original ambition of going to America. We do not find that the Appellant has been at risk. At page B4 of the Respondent's bundle the Appellant stated that his intention was not to stay in England but to try to get to America to stay with his family. We find that that comment amply reflects our conclusions. He stated in answer to section 2.2 that his intention is to go to America to live with his family. At question 2.79 at page B11 he was asked did you come to the UK specifically to seek asylum. He replied yes but as a means of getting to America. We find that had the Appellant genuinely required asylum he would have come to the United Kingdom much sooner or left the Sudan for another country in which he could have claimed asylum much sooner.”
179. Mr Draycott submitted that the Panel had failed to take properly into account the Appellant’s explanation that he had stayed in Sudan because he had to and only left because he feared being deported to Eritrea. The Panel had also not taken account of the background evidence, for example in the COI Report (paras 6.91-6.94) that large numbers of Eritreans took refuge in Sudan and remained there refusing to return to Sudan.
180. We accept the thrust of the COI Report is as Mr Draycott submitted. However, we do not agree that the Panel’s rejection of the Appellant’s evidence is flawed in law. The Appellant’s account in brief was that he wanted to travel to America where his father lived and he remained in Sudan but experienced difficulties in obtaining a visa from the US officials who told him to wait until 2003 (see paras 29- 30, Statement 30 August 2006). The Panel clearly had the Appellant’s evidence in mind in paragraph 22 of its determination. There are several references to passages in the Appellant’s evidence including the evidence about him wishing to go to America and the difficulties he encountered in doing so and his being told he would have to wait until 2003. In fact, he waited until 2004 before leaving Sudan. The Panel considered the Appellant’s account and rejected it. That assessment cannot be characterised as perverse even if another decision-maker could have reached a different conclusion. In reaching its conclusion on this aspect of the Appellant’s evidence, the Panel did not, in our judgment, materially err in law.
181. Mr Draycott also challenged the Panel’s rejection of the Appellant’s evidence that he was not allowed to attend his own wedding in May 1999 but was allowed to spend one week with his wife after the wedding in a military camp. At paragraph 20 of its determination, the Panel said this:
“20. In his statement of 20th April 2004 the appellant referred to “when I went to meet my fiancée I got married to her in the church. I intended to leave the country with my wife at this point however as it was not a convenient time to escape I could not leave.” The Appellant then explained he went back to the military camp after a period of one week as expected by the military officers. In his statement dated 5th May 2004 the Appellant referred to the wedding being done without his presence. No adequate explanation has been put forward by the Appellant as to how the mistaken version of events was given in the first witness statement to which we refer. In the first witness statement in which the Appellant comments on the Home Office refusal letter to be found at page 10 of the main bundle the appellant does not comment on paragraph 31 of the reasons for refusal of asylum letter which deals with this issue. His comments cease in respect of paragraph 20. The Appellant has made a supplementary statement commenting on the reasons for refusal letter. This is to be found at page 1 of the Appellant’s supplementary bundle. Again there is no comment on paragraph 31 of the reasons for refusal of asylum letter. We have set out the Appellant’s explanation in cross-examination. We reject that. The statement of 20th April is detailed and relatively complex. The part relating to the place of the marriage is preceded by the Appellant stating that he went to meet his fiancée. We reject the possibility of double error by the interpreter on one such point alone. We find no adequate explanation has been given for the discrepancy for the same reasons we have set out in the preceding paragraph. We find it intrinsically implausible that the Appellant would have been refused permission to attend the actual wedding but after the marriage ceremony be given permission to stay with his wife for a week after she came to his division. The Appellant states at paragraph 23 of his witness statement at page 6 of the main bundle that he was allowed to stay with her in another camp in the military front. We find it extraordinary that the Appellant made no attempt to escape during this period. He had already been in detention for about six months and suffered considerable ill-treatment which he has related. The Appellant states at paragraph 13 of his witness statement at page 4 of the main bundle that he thought they would kill him. However it was only in May 1999 that he was allowed to stay with his wife. We find it extraordinary that the Appellant should contradict himself within a short period of time with regard to the circumstances of his marriage. We find it intrinsically implausible given the fear which he had already experienced that he did not seek to escape given the circumstances in which he found himself.”
182. Mr Draycott submitted that the Panel erred in law in concluding that it was “intrinsically implausible” that the Appellant would be refused permission to attend his marriage ceremony whilst being allowed to stay with his wife for a week in a camp thereafter. He submitted that the Appellant had adequately explained the contradiction in his evidence about whether he had actually attended the marriage with his wife or whether that had been done so through a proxy. Dealing first with the latter criticism, in our judgement, this does no more than assert a disagreement with the Panel’s conclusion. The Panel identified the discrepancy in the Appellant’s accounts as to whether he was present at his marriage and of the Appellant’s explanation of how this apparent discrepancy had occurred. They rejected his explanation. In our view, as a matter of law they were entitled to do so. On this basis alone they were entitled to reject the Appellant’s evidence in respect of this matter. As regards Mr Draycott’s other criticism of the Panel’s reasoning, he referred us to paragraph 5.49 of the COI Report in which it is stated that:
“Authorities generally permitted three visits per week by family members except for detainees arrested for reasons of national security or for evading national service.”
183. He submitted that this undermined the Panel’s reasoning that it was implausible that the Eritrean authorities would refuse him permission to attend his marriage but at the same time allow him to stay with his wife one week later. It is said that this evidence supports such conjugal visits. It may do so although we note that the exception referred to there, namely those “detainees arrested for reasons of national security or for evading national service”. It may well be, on the Appellant’s account, he fell within this exception in any event. However, whether that is so or not, whilst this supports the availability of family visits it does nothing, in our view, to undermine the Panel’s reasoning which was to contrast the attitude of the Eritrean authorities to the Appellant over a short period of time. The Panel was entitled to take into account that contrast and, in our view, the COI Report does not undermine that reasoning.
184. Mr Draycott criticised the Panel for its view that it was implausible that the Appellant would not take the opportunity to escape if he had been staying with his wife for a week at the military camp at the front. Taken alone, this criticism might have some substance to it. However, it is premised on the assumption that the Panel accepted that this had indeed happened. In reality, the Panel had already concluded on a basis which we consider entirely sustainable in law that they did not accept the Appellant’s evidence concerning his marriage and the subsequent claimed stay with his wife. In truth, this aspect of the Appellant’s evidence is, in the words used in the grounds for review “wholly peripheral to his claim”. Unless it can be said, and we do not think that it can, that the Panel’s conclusion not to accept the Appellant’s evidence in relation to his marriage and stay with his wife was significant in reaching its conclusion on the overall credibility of his evidence in respect of the events central to his claim, Mr Draycott’s attack on this aspect of the Appellant’s evidence, even if well made, takes Mr Draycott’s submissions that the Panel’s overall conclusion on credibility cannot stand no further.
185. Finally, we reject Mr Draycott’s submission that the Panel adopted an attitude of general disbelief such that it failed to give careful consideration to each aspect of the Appellant’s evidence. We are in no doubt, having read the determination, that the Panel considered each aspect of the Appellant’s evidence individually and looked at each incident on its own merits. Indeed, the analysis above of the Panel’s reasoning demonstrates that it did so.
186. Whilst there are some aspects of the Panel’s determination and reasoning that are unsustainable, the ultimate issue for us is whether the Panel’s proper rejection of the crucial parts of the Appellant’s evidence relating to his claim for protection in themselves justified the Panel in rejecting the Appellant’s account. Despite some hesitation, it seems to us that they do. In rejecting the Appellant’s evidence of his detention in 1999, escape and flight to Sudan in May 2000 the Panel was entitled as a matter of law to conclude that the Appellant had failed to establish essential factual building blocks necessary to show that he was at real risk of serious ill-treatment contrary to Article 3 of ECHR. The rejection of these crucial parts of the Appellant’s claim so undermine his credibility and the veracity of his evidence that the Panel was entitled to conclude that he had not made out his case. If he was not to be believed on these central parts of his evidence, the Panel was entitled to conclude as a matter of law that he was not a person of credit such that any part of his account crucial to his claim would be accepted.
187. For these reasons, in respect of the Appellant’s claim to be at risk because he is a military deserter, in our judgment, the Panel’s decision to dismiss his appeal discloses no material error of law.
188. The Appellant has, however, another string to his bow. It was an argument presented to the Panel but rejected by them in paragraph 24 of the determination. It is that the Appellant is entitled to succeed because he illegally exited Eritrea and as such will be perceived as a deserter or draft evader. In paragraph 24 the Panel said this:
“It has been accepted by the Respondent that the Appellant served in the military. Having rejected the Appellant’s account we do not accept that on return he would be perceived as a deserter and an escapee from a military prison. The Appellant therefore fails to come within the criteria set out in KA. We do not accept the Appellant’s account in relation to his departure from the country. We therefore do not conclude that he did not leave on a legal basis.”
189. It was not disputed before us that if an individual establishes that he had left Eritrea illegally there is a real risk that he will suffer serious ill-treatment as a perceived deserter or draft evader on return. That is the position acknowledged in the country guidance case of MA which was approved and followed by the Court of Appeal in GM (Eritrea) and Others v SSHD [2008] EWCA Civ 833. The point in this appeal is, accepting the Panel’s view on the incredulity of the Appellant’s account, did the Panel err in law in concluding that the Appellant had failed to establish that he had left Eritrea illegally. Mr Draycott submitted that the Panel had failed to take account in reaching its conclusion the evidence of the Appellant that he had left Eritrea illegally and that his passport had expired in 1997 (see screening form, at 1.21). Mr Draycott submitted that the Appellant had no reason to lie about this as, the time he gave this information, it was not yet the position that illegal exit per se was considered to create a real risk on return.
190. The difficulty with Mr Draycott’s position is that this issue has to be considered in the context of the Panel having properly rejected totally the credibility of the Appellant’s evidence about his claim. In such circumstances, it was entirely open to the Panel not to accept the evidence he gave about his passport or the nature of his departure from Eritrea. The reasoning of the Panel, in our view, is entirely sustainable and their conclusion on that basis that the Appellant had failed to establish that he had left Eritrea illegally is unassailable.
191. For completeness we should add that we see no other basis upon which the Panel could have reached a finding that the Appellant had left Eritrea illegally. The only evidence accepted by the Panel (indeed accepted by the Respondent at all times) is that the Appellant is Eritrean, that he was thirty-three at the date of the Panel’s decision and that he had, at some time, served in the Eritrean military. It was, of course, part of the Appellant’s case that he had initially been conscripted between 1990 and 1993 before, on his case, being called up for a second time in May 1998. His evidence in respect of the latter was not accepted by the Panel. Any earlier military service does mean, of course, as the background evidence shows, that he was subject to recall at any time. But, as the decision of the Court of Appeal in GM makes clear, that alone does not create a risk on return to an individual. The risk only exists if it is established, albeit on the lower standard applicable in human rights cases, that the Appellant also left Eritrea illegally. As the Tribunal pointed out in MA at paragraph [449]:
“…where a person has come to this country and given what the fact-finder concludes (according to the requisite standard of proof) to be an incredible account of his or her experiences, that person may well fail to show that he or she exited illegally.”
192. That approach was cited and approved by the Court of Appeal in GM (see [13] per Buxton LJ).
193. In MA the Tribunal identified from the substantial body of evidence put before it the situations in which an Eritrean national might obtain an exit visa. At para [348] relying upon the expert evidence of Dr Kibreab the Tribunal said this:
“348. As noted at paragraph 205 above, Dr Kibreab told us that those not affected by National Service and considered as trustworthy by the government, and thus unlikely to have difficulty in obtaining exit visas, comprised Ministers; ex-Ministers; Party Activists; Eritrean expatriates; namely those who could be British citizens working in Eritrea but of Eritrean origin; elderly people over fifty who were forty or over in 1994, those who wanted to go on Haj or visit relatives abroad; government officials; scholarship students (although Dr Kibreab’s evidence was that the government now restricted their movements as many did not return); government employees who attended conferences (although Dr Kibreab maintained this had recently stopped); and relatives of those in power who might arguably obtain exit visas as a result.”
194. The Tribunal’s view was adopted by the Court of Appeal in GM.
195. We accept that there are large numbers of Eritrean citizens who are obliged to leave their country illegally (see MA at [361]). We also accept that a male of military service age – such as the Appellant – would be unlikely to obtain an exit visa unless he came within the limited categories identified in MA (see [357]). In this case, however, there is no evidence one way or another as to the Appellant’s basis for exiting Eritrea. His evidence was not accepted. In these circumstances, in our view, it would be mere speculation to conclude upon what basis he did leave. In the result, there is only one proper conclusion which we can reach, that is that the Appellant has not demonstrated a real risk or reasonable likelihood that he left Eritrea illegally.
196. For all these reasons, we are satisfied that the Panel did not materially err in law in rejecting the Appellant’s claim based upon his illegal exit from Eritrea.
Decision
197. For all the above reasons, we are satisfied that the Panel did not materially err in law in dismissing the Appellant’s appeal on asylum and humanitarian protection grounds or in respect of Article 3 of the ECHR. The Panel’s decision to dismiss the appeal stands.

SENIOR IMMIGRATION JUDGE GRUBB