KS v The Secretary of State for the Home Department

Upper Tribunal

(Immigration and Asylum Chamber)

KS (benefit of the doubt) [2014] UKUT 00552 (IAC)

THE IMMIGRATION ACTS

Heard at Field House	Determination Promulgated
On 2 December 2013, 15 May 2014
	…………………………………

Before

UPPER TRIBUNAL JUDGE STOREY

UPPER TRIBUNAL JUDGE COKER

UPPER TRIBUNAL JUDGE O’CONNOR

Between

KS

(ANONYMITY ORDER MADE)

Appellant

and

THE SECRETARY OF STATE FOR THE HOME DEPARTMENT

Respondent

Representation :

For the Appellant: Mr D Bazini, Counsel instructed by Lawrence & Co Solicitors

For the Respondent: Ms A Everett, Home Office Presenting Officer

(1) In assessing the credibility of an asylum claim, the benefit of the doubt (“TBOD”), as discussed in paragraphs 203 and 204 of the 1979 UNHCR Handbook on Procedures and Criteria for Determining Refugee Status, is not to be regarded as a rule of law. It is a general guideline, expressed in the Handbook in defeasible and contingent terms.

(2) Although the Handbook confines TBOD to the end point of a credibility assessment (“After the applicant has made a genuine effort to substantiate his story”: paragraph 203), TBOD is not, in fact, so limited. Its potential to be used at earlier stages is not, however, to be understood as requiring TBOD to be given to each and every item of evidence, in isolation. What is involved is simply no more than an acceptance that in respect of every asserted fact when there is doubt, the lower standard entails that it should not be rejected and should rather continue to be kept in mind as a possibility at least until the end when the question of risk is posed in relation to the evidence in the round.

(3) Correctly viewed, therefore, TBOD adds nothing of substance to the lower standard of proof, which as construed by the Court of Appeal in Karanakaran v Secretary of State for the Home Department [2000] 3 All ER 449, affords a “positive role for uncertainty”.

(4) The proposition in paragraph 219 of the Handbook, that when assessing the evidence of minors there may need to be a “liberal application of the benefit of the doubt” is also not to be regarded as a rule of law or, indeed, a statement of universal application. As a reminder about what the examiner should bear in mind at the end point of an assessment of credibility, the proposition adds nothing of substance to the lower standard of proof. If, for example, an applicant possesses the same maturity as an adult, it may not be appropriate to resort to a liberal application of TBOD.

(5) Article 4(5) of the Qualification Directive is confined to setting out the conditions under which there will be no need for corroboration or “confirmation” of evidence. Although (unlike the Handbook) Article 4(5) does set out conditions that are rules of law, properly read, it is not to be compared with the scope of TBOD as described above.

DETERMINATION AND REASONS

Discussion

Different understandings: two opposite approaches

54. It is clear from our earlier survey that significant differences exist over the understanding of TBOD in the asylum field. At the risk of caricature, it is possible to discern two approaches at opposite ends of a spectrum. At one extreme, there is what we shall call the “pervasive approach” which construes the notion as one which must be applied to assessment of each and every item of evidence. This appears to be the approach advocated by La Forest J in the Canadian Supreme Court of Chan , at least as his approach was portrayed by Major J at [142]: see above paragraphs 9-10. At the other extreme, there is the “end-point” approach, which understands the notion to be one that only applies at the end-point of assessment after the examiner has established certain facts about the applicant’s story and only when there is doubt about certain remaining aspects of that story. The doubt arises because certain essential material facts are not confirmed (or corroborated) or supported by other evidence. As articulated in the discussion of the rule in the same Chan case, this latter approach is seen to be typified by that set out in the 1979 UNHCR Handbook. One can even find examples where these two contrasting positions are asserted at the same time, as is done by the respondent in her current guidance on “Considering asylum claims and assessing credibility” at paragraph 4.3.4: see above paragraphs 21, 45.

55. Clearly there are variants of both these approaches. We have identified as the “holistic approach” one in which TBOD is given whenever doubt arises. Certainly at least for some commentators Article 4(5) of the QD is akin to but not on all fours with the UNHCR Handbook’s “end-point” approach. The 2013 UNHCR CREDO study records that in its review of case files in the UK the rule is seen to apply to intermediate stages of credibility assessment or more precisely, [this review] “indicated that some decision-makers may apply the principle of the benefit of the doubt mid-way through the credibility assessment, before considering the evidence in the round” (paragraph 2.4.3.2, p. 233).

56. How are we to resolve these different understandings?

Usual meaning

57. In common idiom, giving someone TBOD is “to decide that you will believe someone, even though you are not sure that what the person is saying is true” (Cambridge Academic Content dictionary). Insofar as the term applies to assessment of evidence in a legal context, a precise definition of the phrase is elusive but it cannot mean much more than the following. For TBOD to be given something about the evidence must be in doubt. For evidence to be in doubt something about it must not be certain. If it was certain, there would be no need to give any benefit. Benefit being given in the face of doubt indicates that the examiner is not entirely convinced whether an asserted fact is true or not, and so gives the subject the benefit of the most positive outcome they can, which is to accept the facts narrated. Thus stated, it is a rule or practice that can be applied at any time during an inquiry into some asserted fact(s). It can be applied to the first, or last asserted fact(s) in a series of asserted facts, or any asserted fact(s) in between. But recourse to the usual meaning does not necessarily help in an area of law which is in many ways, sui generis, and certainly (as emphasised by the Court of Appeal in Karanakaran ) unlike ordinary civil litigation.

Rationale

58. Looking at recourse made to the rule in asylum law and also in human rights law dealing with the non-refoulement context – and notwithstanding the different understandings alluded to above - there is considerable consensus about the underlying rationale of TBOD notion, which is twofold. First it reflects what A. Grahl Madsen described as “the well-known fact that a person who claims to be a refugee may have difficulties in proving his allegations”: see e.g. A. Grahl-Madsen, The Status of Refugees in International Law , 1966, 145-146); the ECtHR in R.C. v Sweden . We shall term this the “ameliorative” aspect.

59. Second, the notion helps insure against the particularly grave consequences of “getting it wrong” in a context in which the stakes are high: sending back persons wrongly refused international protection status could result in their death or ill-treatment: see Karanakaran (Brooke LJ at [44], Sedley LJ at [17]). We shall term this the “precautionary” aspect.

60. It will be observed as well, that very similar reasons are also given to explain the adoption of a lower standard of proof or for there being a shared burden of proof or a shared duty of cooperation on the applicant and respondent in certain contexts.

Not a rule of law

61. In order to arrive at a proper analysis of the scope and application of TBOD notion, it is first of all important to understand what it is not. It is not to be considered as a rule of law. As stated by Laws LJ in MD (Guinea ) at [12] (in relation to the “liberal application of the benefit of the doubt” notion):

“… it is of particular importance that the plethora of guidance coming from many sources is not to be degraded into a set of concrete rules, departure from any one of which then falls to be characterised as an error of law. It is not for example a rule of law that a child's evidence should be accorded "a liberal application of the benefit of the doubt", a phrase appearing in some of the guidance documents. That said, the phrase represents or points to an approach which in some cases it may be very useful to have in mind.”

62. Significantly, even the 1979 UNHCR Handbook recognises that the notion is a defeasible one: paragraph 203 states that it is “frequently”, not invariably “necessary to give the applicant the benefit of the doubt”. Likewise the “liberal application of the benefit of the doubt” notion set out at paragraph 219 is expressed as being one which it “may” be necessary to apply to the known circumstances of a minor. Similarly, the ECtHR states that giving TBOD to asylum seekers when it comes to assessing credibility is something which it is “frequently”, not always, necessary to apply to their evidence: see R.C. v Sweden , supra paragraph 7.

Relationship to the burden of proof

63. Manifestly, at least as concerns the position in UK law, TBOD notion cannot as such affect the burden of proof, which remains on the applicant: see Elias LJ in HK (Afghanistan ) at [34]. In passing, we would observe however that the duty imposed on the Member State by Article 4(1) QD) - and on decision makers in the UK by paragraph 339I of the Immigration Rules - to assess the relevant elements of an application “[i]n cooperation with the applicant …” (emphasis added) must at least have an ameliorative effect in respect of an applicant being able to discharge that burden.

Relationship to the standard of proof

64. Understood in terms of its underlying rationale, it is relatively easy to regard TBOD notion as an aspect of the lower standard of proof applied in the field of asylum (and asylum-related) law. If a key feature of the lower standard of proof is, in the words of Brooke LJ in Karanakaran , affording “a positive role to uncertainty”, then at least when confronted with doubts about an applicant’s account, the decision-maker must consider resolving at least some of them in his or her favour. Because the notion only applies when there is doubt, it clearly cannot affect the standard of proof itself. It can only regulate what response is to be given consonant with that standard where uncertainty or doubt arises.

65. At the same time (as already noted), on the same logic, the notion cannot be a “pervasive” one to be applied to each and every item of evidence. To apply it thus would be to ignore that there are certain asserted facts which can and should be accepted or rejected without reference to the notion of TBOD. Thus in the U.S. case law it has been emphasised that if an applicant’s account lacks credibility in key respects, then the notion has no application: see e.g. Sukwanputra v Gonzalez , 434 F.3d 627, 634-35 (3 ^rd Cir, 2006). The point is explained more fully in the 2013 UNHCR CREDO study a 4.2 (p.247) with emphases added in italics:

“An asserted fact may be accepted because it is sufficiently detailed, internally consistent, and consistent with information provided by family members and witnesses, consistent with available specific and general objective COI, and plausible when considered in light of the applicant’s individual and contextual circumstances. Such facts may be accepted without reference to the principle of the benefit of the doubt.

An asserted fact may be rejected because, when taking into account the reasonableness of the explanations provided by the applicant with regard to the potentially adverse credibility findings and the applicant’s individual and contextual circumstances, the applicant’s statements about that fact are not sufficiently detailed, consistent, and plausible, and/or are contradicted by other reliable, objective, and time appropriate evidence. Again, such facts may be rejected without reference to the principle of the doubt because the principle cannot be applied to remedy what is clearly not credible based on all the available evidence.”

66. Indeed there is a more fundamental reason why a ”pervasive” approach should be discouraged. Even considered as one aspect of the lower standard of proof, the latter is ultimately about the proper approach to establishing proof of a risk and consideration of risk can only properly be made at the end-point. As expressed in slightly different language by Kirby J in Wu Shan Liang (1996) 185 CLR 259 at [31] (cited with approval by Brooke LJ and Sedley LJ in Karanakaran at [78] and [17] respectively):

"Ultimately the question is whether the delegate [i.e. the decision-maker] allowed her mind to consider all the relevant possibilities by looking back at the entirety of the material placed before her and considering it against a test of what the "real", as distinct from fanciful, "chances" would bring if the applicant were returned to China."

67. Sedley LJ returned to the same theme in SR (Iran) v Secretary of State for the Home Department [2007] EWCA Civ 460 – a case in which Mr Bazini was also Counsel for the appellant in the following terms:

“6. Mr Bazini's argument on this issue seems to me to make the basic mistake of seeking to import the ultimate question of risk into the evaluation of each piece of evidence. For example, he argues that the AIT, had it approached its task properly in §41, might well have accepted "that there was a real risk of [A's] bearing witness" in Iran. On this premise he criticises the finding that A was not likely to bear witness to her friends in Iran as setting too high a standard of proof.

7. It is evidently necessary to say something fairly basic about the key exercise of fact-finding and risk evaluation in asylum and human rights cases. In Karanakaran [2000] 3 All ER 449 this court gave guidance which included, at 479, the following:

"No probabilistic cut-off operates here: everything capable of having a bearing has to be given the weight, great or little, due to it … The facts, so far as they can established, are signposts on the road to a conclusion on the issues; they are not themselves conclusions."

8. Applying this approach to the present case, it was the AIT's task, first, to discard any evidence judged to be of no value at all: here, for example, the account which, for better or for worse, had been disbelieved on an earlier hearing. For the rest, the AIT had to take each element of evidence into account for what it was worth. Some of the evidence was worth a good deal to the appellant: for example the in-country evidence adopted in §38 and expanded thereafter about how Christians are marginalised and subjected to discrimination in Iran. Some of it, in the AIT's judgment, was less compelling: for example that the appellant would expose herself to persecution by evangelising.

9. There is nothing wrong with the differential levels of proof or disproof of primary facts found by the tribunal. In §41, for example, they find it likely that the appellant would continue with Christian communion in Iran and that this would bring her to the attention of the authorities, but they do not accept that she would bear witness to her friends. In §46 they find that the fact that her parents are pensioners does not necessarily mean that they would be unable to protect her economically from persecution. The law does not demand, at least in this field, that each finding of fact, whatever its degree of certainty or uncertainty, be fitted into a single matrix of risk. The fact-finder's task is, to the extent made possible by the evidence, to find facts, and some facts are more certain than others. It would have been as unjust to the appellant to treat as mere possibilities things which, on the AIT's findings, were highly likely as it would have been to the respondent to treat possibilities of hardship as probabilities.

10. The critical adjudicative task is to assemble these findings into an evaluation which answers the question posed by law. In asylum and human rights claims, that is the question of real risk, and it is at the point of decision and not sooner that it arises.”

68. It is clear from this analysis that it is incorrect for fact-finders in asylum cases to apply the lower standard to each and every piece of evidence except insofar as they are seeking to ensure that no asserted fact or item of evidence should be excluded unless there is certainty or finality about its value or lack of it: this brings us back to what Brooke LJ described as “a positive role for uncertainty”.

69. But whilst in this holistic but less open-ended way the rule may be said to inhere in the (lower) standard of proof, it adds nothing and only risks confusion to describe it as amounting to a TBOD “principle” or “rule” or “approach” or “basic concept (Footnote: ³) ”. It is not a rule of law or universally applicable principle, whereas the standard of proof to be applied in the asylum field is a fundamental tenet.

70. Equally, even on the above analysis, it is difficult at least at first blush to see that the notion can be confined in the way in which it appears to be in the 1979 UNHCR Handbook or in the 2013 UNHCR CREDO analysis reformulation according to both of which the notion only comes into play at the end point of the credibility assessment, after the decision-maker has sought to accept or reject what asserted facts he can. (The characterisation of this as an “end point” approach is clearly stated in paragraph 4.1 (pp.246-7) of the CREDO study quoted at paragraph 14 above)

71. At least when construed as a means of ensuring a “positive role for uncertainty”, it is hard to see why TBOD notion should not potentially apply even at an early stage of the credibility assessment, e.g. when an applicant is stating basic particulars (or “biodata”) concerning where they live or what is their ethnic origin. If at an initial or intermediary stage there is uncertainty such that some item of evidence points one way, whilst another item points the other way, surely this is precisely the type of context in which the decision-maker should (in this limited sense) give TBOD.

72. Taken at face value, then, it is difficult to disagree with Mr Bazini that the 1979 UNHCR Handbook “end-point” understanding of the rule is too restrictive.

73. On reflection, however, we agree with Ms Everett (and consider that Mr Bazini at one point acknowledged as much) that it is possible to reconcile the two approaches. Given that the “pervasive approach” overstates the scope of the rule’s potential application (since if there is no doubt, there is no benefit and the lower standard of proof is essentially about how to approach proof of the ultimate question of risk) what is involved in this wider-ranging (albeit not untrammelled) holistic approach is simply no more than an acceptance that in respect of every asserted fact when there is doubt, the lower standard entails that it should not be rejected and should rather continue to be kept in mind as a possibility at least until the end when the question or risk is posed in relation to the evidence considered in the round. In our judgment the fact that the Court of Appeal in Karanakaran did not describe its espousal of a positive role for uncertainty as a TBOD rule reflects this fact. Their lordships in Karanakaran built closely on Australian case law. Encapsulating that, in the Australian case of Harjit Singh Randhawa v Milgea [1994] 124 ALR 265, Beaumont J stated that decision-makers should take a “liberal approach” to the assessment of evidence. In Rajalingam (Aus.FFC,1999) the same Court stated at [240] that:

“[It] is not always possible for the decision-maker to be satisfied as to whether alleged past events have occurred with certainty or even confidence. When the [decision-maker] is uncertain as to whether an alleged event occurs, or finds that, although the probabilities are against it, the event might have occurred, it may be necessary to take into account the possibility that the event took place in considering the ultimate question”.

74. We also take this to be the underlying reason why Hathaway and Foster in the new edition of Law of Refugee Status , 2014, after an analysis citing the Australian case law with approval, conclude that (what they refer to as) “the principle” of TBOD really adds nothing of substance to the lower standard of proof or what they call the“intentionally low threshold of the test of well-founded fear”.

75. We still consider, however, that it is legitimate to continue identifying and applying the rule of TBOD in accordance with the “end-point” formulation given in the 1979 UNHCR Handbook so long as it is understood as no more than useful particularisation of the lower standard of proof which, as we have seen, affords a positive role for uncertainty when the decision-maker is confronted with an applicant who has succeeded in persuading the decision-maker of certain asserted facts but cannot dispel doubts about all of them and a conclusion has to be reached as to the overall assessment. Whereas we have earlier concluded that (i) the “pervasive” approach to the notion is incorrect (because it cannot be applied to each and every item of evidence); and (ii) the “holistic approach” does not add anything of substance to the lower standard of proof, we think that in the context identified in the 1979 Handbook, reference to the notion does help clarify the task in hand and to make clear that before any “signing-off” the evidence has to be considered overall. Such an approach also acts as a safeguard against any premature evaluation of the ultimate question of risk of the type abjured by Sedley LJ in SR (Iran ).

76. To the extent that it can be argued that the “end-point” approach set out in the 1979 UNHCR Handbook is too restrictive, it seems to us that the answer is that just because the rule is only (or should only be) given specific articulation at the “end-point” does not mean that an asylum applicant has not had the benefit earlier on of a positive role for uncertainty.

77. To the extent that it could be argued that to apply the “end-point” formulation of the rule is tantamount to giving an applicant (impermissibly) TBOD twice over, our answer is that since the underlying premise of the “end-point” formulation is that earlier on in the assessment the decision-maker has afforded throughout a positive role for uncertainty (which we take to be the thrust of paragraphs 195-202 of the UNHCR Handbook), to apply it in this way simply ensures it is not overlooked at one particular (and particularly important) stage.

The TBOD in Strasbourg jurisprudence

78. We find it unnecessary in this decision to seek to interpret the precise import of the approach to TBOD set out by the ECtHR in its jurisprudence relating to Article 3 in the context of assessing risk on return, but we would make this observation. Whilst that approach is clearly wider than an “end-point” approach, we are not convinced that it is as holistic or pervasive as Mr Bazini submits, since the Court appears to envisage that in certain circumstances the notion is disapplied “when information is presented which gives strong reasons to question the veracity of an asylum seeker’s submissions”: see I v Sweden , 61204/09 [2013] ECtHR 813, at [50] (supra paragraph 8. ( Insofar as it might be suggested that in R.C. v Sweden the Court was envisaging that in certain circumstances the burden of proof shifts from the applicant to the state, we would regard ourselves as bound to follow domestic authority to the contrary: see PJ (Sri Lanka ) [2014] EWCA Civ 1011, approving NA (UT rule 45: Singh v Belgium ) Iran [2014] UKUT 205 (IAC) and MJ (Singh v Belgium: Tanveer Ahmed unaffected ) Afghanistan [2013] UKUT 254 (IAC ).)

Article 4(5) of the Qualification Directive (QD)

79. Perhaps unsurprisingly given the dearth of case law, neither party was able to give us much assistance in relation to the question as to whether Article 4(5) QD has replaced or modified the UNHCR Handbook’s formulation of TBOD notion. Since our hearings dealing with the appellant’s case some light has been shed on Article 4(5) by Advocate General Sharpston in her Opinion in Joined Cases C-148/13, C-149/13 and C-150/13, A, B and C , 17 July 2014. At [50] she observes that in general terms neither the Qualification Directive nor the Procedures Directive (2005/85/EC) make specific provision for the manner in which an applicant’s credibility is to be assessed and that “the general position is that, in the absence of EU rules on a subject, it is for the domestic legal system of each Member State to determine the procedural conditions governing actions at law intended to ensure the protection conferred by EU law.” Earlier, at [42], she observed that:

“42.      Article 4(1) allows Member States to place the onus upon applicants ‘… to submit as soon as possible all elements needed to substantiate the application for international protection …’.  That provision also places a positive duty on Member States to act in cooperation with the applicant to assess the relevant elements of his application. The assessment should be carried out on an individual basis and should include taking into account the applicant’s individual position and personal circumstances.  Article 4(5) of the Qualification Directive acknowledges that an applicant may not always be able to substantiate his claim with documentary or other evidence. Such evidence is therefore not required where the cumulative conditions of Article 4(5)(a) to (e) are met.”

80. The Advocate General goes on to emphasise at [56]-[57] that “ it does not follow from the lack of any express wording in the Directive regulating Member States’ discretion regarding the practices or methods for assessing an applicant’s credibility, that EU law places no limits on that discretion.” She highlights the role of the Charter:

“57.      The Charter provides overarching standards that must be applied in the implementation of any directive. The Qualification Directive harmonises by introducing minimum standards for obtaining refugee status within the European Union.  It would undermine the CEAS, in particular the Dublin system, if Member States were to apply widely divergent practices when assessing such applications. It would be undesirable if the differences in its implementation led to applications being more likely to succeed in one jurisdiction than in another because the evidentiary requirements were easier to satisfy. “

81. In light of this Opinion and other academic commentary it can be discerned that the effect of Article 4(5) (for those Member States who opt to apply the principle according to which it is the duty of the applicant to substantiate the application for international protection) is limited to the situation where aspects of the applicant’s statements are not supported by documentary or other evidence (i.e. a situation where there is a lack of corroboration). By Article 4(5) the applicant in such a situation will not need to provide confirmation (corroboration) when the conditions enumerated under (5) are cumulatively met (Footnote: ⁴) . These conditions relate to genuine effort (4(5)(a)), satisfactory explanation (4.5(b)), plausibility (4(5)(c), earliest possible application (4(5)(d) and general credibility (4(5)(e)).

82. As such, does Article 4(5) enunciate a TBOD rule? We have already referred to the position taken by the respondent in the APIs which is that this provision does enunciate such a rule (see above paragraphs 18,21). That is lent some degree of support by the observation made by Hailbronner in EU Immigration and Asylum Law-Commentary , 2010, at p.1030 that the text of Article 4(5) was based in part on the UNHCR Handbook.

83. However, stating that Article 4(5) is based “in part” on the HCR Handbook does not demonstrate that this provision codifies such a rule and the idea that it does so is not shared by the most in-depth academic commentary so far, that by Gregor Noll, “Evidentiary assessment and the EU qualification directive”, UNHCR Working Paper no. 117 in New Issues in Refugee Research , June 2005. Describing Article 4(5) as comprising overall “a qualified alleviating evidentiary rule”, Noll cautions at p.12:

“The wording of [this rule] signifies a heavier burden for the applicant than the analogue principle of “benefit of the doubt”, as this is described in UNHCR's handbook. In addition to the requirements given in the handbook, article 4.5 places at least one additional condition on the applicant', s/he is required to have applied for international protection as early as possible unless s/he can provide good reasons for not doing so”.

84. Of course, had we accepted that Article 4(5) of the QD is to be read as codifying a TBOD rule, then we would have to revise our earlier statement that this notion is not a rule of law. That is because, in the UK, Article 4(5) has been implemented by paragraph 339L of the Immigration Rules (see above, paragraph 20) which for most purposes do have the force of law.

85. But (like Noll) we think it unwise to refer to Article 4(5) as reflecting or encapsulating a TBOD rule. In part that is because we think that the provision seeks primarily to identify one particular application of the lower standard of proof (which in our view affords a positive role for uncertainty). In part this relates to the point we have already emphasised, namely that this provision is, limited to cases where there is a lack of corroboration: viz., “where aspects of the applicant’s statements are not supported by documentary or other evidence” (chapeau to Article 4(5)); that on its face is more limited in scope than the notion as set out in paragraph 204 of the UNHCR Handbook (where the precondition is that “all available evidence has been obtained and checked and when the examiner is satisfied as to the applicant’s general credibility”). In part is because whereas the notion is formulated in the UNHCR Handbook in defeasible and contingent terms, Article 4(5) sets out in mandatory terms (Footnote: ⁵) four conditions which have to be cumulatively met in order for an applicant not to be required to substantiate his application. As such we think Article 4(5) (and paragraph 339L of the Rules) should be applied as it is without any gloss or reference to the UNHCR Handbook “analogue”. We understand UNHCR’s concern, as expressed in the 2013 CREDO study, that if not read in the light of the UNHCR Handbook, Article 4(5) could operate as a policy for withholding TBOD from asylum-seekers; but we consider that danger to be offset by (i) the fact that the ambit of Article 4(5) is limited to cases of non-corroboration/confirmation; and (ii) the fact that applicants benefit throughout the assessment of credibility from the lower standard of proof, which in the terminology of Karanakaran , ensures that the entirety of that assessment affords a positive role for uncertainty.

TBOD and children

86. We are now in a position to address the main specific issue in this case, which concerns whether a decision-maker must make a “liberal application” of TBOD notion when it comes to assessing the credibility of a minor. In essence Mr Bazini’s central contention on this issue was that, in general, when assessing the credibility of minors the “principle” of liberal application of TBOD should be applied, and certainly cannot be disapplied on the basis of a child’s maturity. Ms Everett disagreed.

87. We deal first with the notion of a liberal application of TBOD as it is expressed in paragraph 219 of the 1979 UNHCR Handbook.

88. It is clear from the language of paragraphs 213-219 of this Handbook (which deal with assessment of the credibility of minors and paragraph 219 (which expressly refers to “a liberal application of the benefit of the doubt”) that the notion has a broader scope than the “end-point” analysis set out in paragraph 203. Notably, paragraph 217 states that where the minor has not reached a sufficient degree of maturity to make it possible to establish well-founded fear in the same way as for an adult, “it may be necessary to have greater regard to certain objective factors” and paragraph 219 states that:

“219. If the will of the parents cannot be ascertained or if such will is in doubt or in conflict with the will of the child, then the examiner, in cooperation with the experts assisting him, will have to come to a decision as to the well-foundedness of the minor’s fear on the basis of the known circumstances, which may call for a liberal application of the benefit of the doubt.”

89. Both of these paragraphs envisage that (a liberal application of) TBOD may need to be given to assessment of the credibility of a child at every stage; certainly there is no confinement to just an “end-point”. This might appear to suggest that in relation to children TBOD notion has a conceptual life of its own.

90. By taking trouble to first set out the learning in the asylum context on the notion of TBOD in general, we hope it will be apparent that (and we do not understand either party to submit otherwise), the “liberal application” concept can only be adjectival and cannot affect the underlying nature of the notion. If, as we have held, TBOD is neither a legal rule nor a principle nor a notion to be applied invariably, it cannot be converted into one by a more liberal application of it. (The fact that is not a rule to be applied invariably is clear in any event from the wording of the Handbook which itself treats it as a discretionary rule: paragraph 219 states that a minor’s known circumstances “may” call for a liberal application of the notion.) Accordingly, except insofar as it operates as a useful reminder of what a decision-maker should do at the “end-point” of the assessment of the credibility of a minor, the notion of a liberal application of TBOD is best understood simply as an expression of how the lower standard of proof is to be applied in making such an assessment. Application of that standard when the subject-matter is the evidence of a child must take account of the fact that the applicant is a child and that children in general are vulnerable in a way most adults are not.

91. There are, however, arguably two extra dimensions to this notion that arise in the UK context. The first possible dimension brings us back to paragraph 351 of the Immigration Rules which, we recall, provides that:

“A person of any age may qualify for refugee status under the Convention and the criteria in paragraph 334 [dealing with Grant of Asylum] applied to all cases. However, account should be taken of the applicant’s maturity and in assessing the claim of a child more weight should be given to objective indications of risk than to the child’s state of mind and understanding of his situation. An asylum application made on behalf of the child should not be refused only because the child is too young to understand the situation or have formed a well-founded fear of persecution. Close attention should be given to the welfare of the child at all times.”

92. The second possible dimension consists of the Asylum Policy Guidance (APG) whose key provisions we also cited earlier but for ease of reference we set out again here. They state that the following principles to be applied in the determination of a claim to asylum by a child:

“(1) more weight must be given to objective indications of risk than to the child’s state of mind;

(2) other factors to consider might include: documentary evidence, objective country evidence;

(3) a case owner must not draw an adverse credibility interference from omissions in the child’s knowledge if it is likely that their age or maturity is a factor or if there are logical or other reasons for those omissions;

(4) the benefit of the doubt will need to be applied more generously when dealing with a child particularly where a child is unable to provide detail on a particular element of their claim.”

93. As to the first possible extra dimension, we note that paragraph 351 does not articulate any TBOD notion and quintessentially its contents primarily convey what is necessary in order for an examiner to ensure sensitive application of the lower standard of proof in respect of the evidence of a child. As to the second possible dimension, we straightaway acknowledge that if the APG provisions are more generous than the notion as expressed in the 1979 Handbook, then it would be not in accordance with the law for a decision-maker to fail to apply a more generous policy: see Abdi v Secretary of State for the Home Department [1995] EWCA Civ 27. However, first of all the APG places reliance on the guidelines set out in the 1979 Handbook and does not purport to go beyond them; and secondly, the factors that are highlighted as ones the examiner should bear in mind are, again, best understood as expressing what is entailed in applying the lower standard of proof to the assessment of the evidence of a child. Accordingly, we consider on analysis that neither paragraph 351 of the Rules nor the APG provisions relating to the evidence of children alter the general position we have reached in relation to the TBOD notion in general.

94. However, whether understood in the way we have just set out or (as the parties contended) as a distinct “principle”, we still need to engage with Mr Bazini’s underlying argument that, in general, when assessing the credibility of minors liberal application of TBOD should be given, and certainly cannot be disapplied on the basis of a child’s maturity. Indeed it was the failure to appreciate this point, he said, that led the First-tier Tribunal judge into one of a number of legal errors. He submits that valid reasons for disapplying the notion should only have to do with the nature of the applicant’s evidence (there being no reason to apply it, for example, if the applicant has given a credible story in all respects or has failed to establish any elements of his claim). Such reasons should not or cannot relate to characteristics of an individual applicant. According to Mr Bazini, characteristics such as age, education, intelligence, health etc may have a bearing on credibility but do not in themselves provide a basis for disapplying TBOD.

95. Even leaving aside whether (as we have held) the notion of TBOD adds nothing of substance to the lower standard of proof, we are not persuaded that Mr Bazini’s general proposition as regards the liberal application of TBOD to children is tenable. The UNHCR Handbook itself emphasises more than once that in asking whether a minor may qualify for refugee status that must be “determined in the first instance according to the degree of his mental development and maturity”: see paragraph 214. Further, the Handbook only describes it as being necessary to have greater regard to certain objective factors – one of the main consequences of applying a liberal TBOD - if “the minor has not reached a sufficient degree of maturity to make it possible to establish a well-founded fear in the same way as for an adult”: see paragraph 217.

96. To our understanding that indicates that for the Handbook the liberal application of TBOD, which is expressed in defeasible terms anyway, may not necessarily be given if the applicant has equivalent maturity to an adult.

97. Mr Bazini sought to argue that if the FtT judge’s approach (of considering that maturity justified disapplying TBOD notion) were correct, it would call into question the “very principle of the benefit of the doubt” applied to asylum-seekers generally. Even again assuming this notion amounts to a distinct rule, we cannot agree. For a judge to proceed to assess the evidence of a minor without any indication that he or she has considered the modified approach enjoined by paragraph 351 may often result in an error of law, because under that provision more weight must be given to objective evidence. But that is quite different from the situation where a judge considers whether either paragraph 351 considerations or the (liberal application) of TBOD should be given and decides not to give it. We disagree with Mr Bazini’s reasoning for two main reasons. First of all, the fact that the notion per se is a defeasible not a categorical one entails that it is capable of being disapplied in certain circumstances. Second, contrary to what Mr Bazini asserts, one does when assessing the evidence even of adult asylum-seekers, distinguish between for example more or less intelligent or educated individuals. Indeed, failure to take account of such differences may well infringe the rule set out in Article 4(3)(c) of the Qualification Directive (also expressed in paras 41-43 of the UNHCR Handbook), that assessment of a claim for international protection includes taking into account “the individual position and personal circumstances of the applicant including factors such as background, gender and age…”

An adult applicant’s evidence about events when he was a child

98. In his skeleton argument at [11] Mr Bazini stated that:

“In respect of adults who are relying on events that took place at a time when they were minors, it is submitted that insofar as they rely on any evidence given in interviews and statements etc when they are still minors, the more liberal benefit of the doubt principle should be applied. One could not expect the same principle to be applied to evidence generally given after they had attained the age of majority. However, if an adult (e.g. a 20 year old) is giving evidence that he experienced/observed when he was a minor it has to be noted that memory and understanding of these events will have been as a minor and therefore this specific evidence should have the liberal benefit of the doubt approach applied to it. It would be irrational to accept the evidence of a 15 year old in respect of a particular account of events but reject that some account given three years later, once he had turned 18, unless it could clearly be shown that his understanding and knowledge of the events should have become greater or more sophisticated over the three year period etc.”

99. Whilst we have declined to follow Mr Bazini’s view that the liberal application of TBOD is a cardinal principle, we would concur at least that a child-sensitive application of the lower standard of proof may still need to be given to persons if they are recounting relevant events that took place at a time when they were minors or were even younger minors

100. It is also necessary to keep in mind that there is not necessarily a “bright-line” for assessment of evidence in the context of asylum between the stage at which a person is a minor and the stage at which he is an adult: see e.g. KA (Afghanistan) and others v Secretary of State for the Home Department [2012] EWCA Civ 1014 at [18].

Joint Presidential Guidance Note No.2 of 2010

101. For completeness (in order to address a further question we posed to the parties), we agree with the parties that the Joint Presidential Guidance Note No.2 of 2010 (which is based on the Senior President’s Practice Direction First-tier and Upper Tribunal Child, Vulnerable Adult and Sensitive Witnesses October 2008) is neutral as to the issue of whether applicants are entitled to TBOD, although they do reinforce the important point (also set out in paragraph 351 of the Immigration Rules and in Asylum Policy Instructions) that when assessing the evidence of children (and other vulnerable persons) certain factors may affect such a person’s ability to give coherent evidence.

Summary of main conclusions on the two main issues

102. Drawing together the points we have made in the course of our analysis so far we conclude:

a) In the context of assessing the credibility of an asylum claim, the notion of TBOD as set out in paragraphs 203-204 of the 1979 UNHCR Handbook is not a rule of law and even as a general guideline is expressed in defeasible and contingent terms.

b) In its usual meaning the notion can potentially apply at any stage of the assessment of credibility and is not limited, as it is in paragraphs 203-204 of the Handbook, to consideration of its “end-point”; but even so it cannot to be applied to each and every item of evidence, irrespective of whether there is doubt about it.

c) Seen as a notion applying holistically TBOD adds nothing of substance to the lower standard of proof, which, as construed by Brooke and Walker and Sedley LLJ in Karanakaran v Secretary of State [2000] 3 All ER, 449-480, affords a “positive role for uncertainty”.

d) Nevertheless, as formulated within the limited context set out in paragraphs 203-204 of the 1979 UNHCR Handbook, the notion serves as a useful reminder of one of the most important things a decision-maker must do at the “end-point” of any credibility assessment. To apply it within the limited context set out in these paragraphs of the Handbook does not entail in effect giving an applicant TBOD twice over; it simply particularises in a specific way what it means to apply the lower standard of proof when certain, but not all, aspects of an applicant’s claim have been established.

e) Whilst the notion of TBOD as articulated in the 1979 UNHCR Handbook was one source for the drafting of Article 4(5) of the Qualification Directive, and whilst there is some overlap with this provision, the latter, which is confined in any event to setting out the conditions under which the need for corroboration or “confirmation” can be dispensed with, must be applied in its own terms. Unlike the Handbook Guidelines, it sets out conditions that are rules of law.

f) The notion as set out in paragraph 219 of the UNHCR Handbook that when assessing the evidence of minors there may need to be a “liberal application of the benefit of the doubt” is, like its parent notion, neither a rule of law not a rule of universal application and, save as a reminder about what the examiner should bear in mind at the “end-point” of assessment of credibility, adds nothing of substance to the lower standard of proof. Further, depending on individual circumstances of the case, it may be a reason not to apply the notion if an applicant possesses the same maturity as an adult.

Decision on error of law

103. In light of the above, we turn to consider whether the FtT judge erred in law. We observe that in the appellant’s case he was not someone who had failed to submit any corroborative evidence; he had submitted his father’s ID card and a photograph of his father with the Hezb-i-Islami leader. We recall to mind that Mr Bazini’s submissions advanced five grounds: see above paragraph 34.

104. In respect of the first ground we would reject Mr Bazini’s argument that the mere fact of declining to apply the liberal benefit of doubt rule to the appellant amounted to an error of law. As explained above, the parent rule of the TBOD is not to be understood as a legal rule nor as a principle and even in its classic formulation it is not understood as requiring invariable application. The same cannot be any less true in relation to the concept of a liberal application of this notion which is also, in any event, expressed in defeasible terms.

105. We would, however, agree with Mr Bazini that some of the reasons given by the judge for disapplying (the liberal application of) the notion, even when understood as a rule ensuring a positive role for uncertainty, were tenuous. If the judge wished to construe the appellant’s journey to journey to the UK as evidence of maturity, then he needed to explain why he did not accept that the appellant was being chaperoned by agents or had left all the arrangements to them. If the judge wished to rely in this regard on the comprehensivity of the appellant’s account, then he needed to explain why at [87] he held against him the fact that he was not able to give a detailed picture of his brother’s activities.

106. These failures of explanation were of particular importance because – irrespective of whether he needed to apply a benefit of the doubt rule - the judge should have started from the position that (at least in relation to the appellant’s recounting of events when he was a minor) paragraph 351 had potential application and hence that “in assessing the claim of a child more weight should be given to objective indications of risk than to the child’s state of mind and understanding of his situation”. That in turn would have enhanced the potential of the expert report of Dr Giustozzi (a report which drew on various country of origin reports) to shed light on the appellant’s case. Unlike TBOD notion set out in the UNHCR Handbook paragraph 351 identifies obligations on the decision maker which have the force of law.

107. From the above it will be apparent that the judge nowhere addressed the issue of to what extent his assessment should take into account the accepted fact that the key events recounted by the appellant had occurred when he was a minor: for example, on the judge’s account when the appellant said his family home had been visited by the authorities on two previous occasions, at most he would have been no more than 14-16. The judge nowhere suggested he thought that when aged 14-16 the appellant was mature for his age; hence no reason whatsoever is given for not affording a child-sensitive approach to this aspect of his claim (so as to ask whether recollection of things that had happened when the appellant was a child might have been less reliable than if they had happened when he was an adult).

108. We also consider that the determination was vitiated by legal error through a failure to engage sufficiently with the evidence of the expert witness, Dr Giustozzi. The judge does refer to his report when dealing with the issue of general risk on return: see [94] and in the first two sentences of [88]. In the remainder of [89] the judge notes:

“According to the evidence, the appellant’s father was well-regarded as a Hezb-i-Islami commander in the area while his elder brother started to dress as a Hezb-i-Islami member and to carry a Kalashnikov rifle a year before he left home. The appellant says he became aware that people were reporting his father and brother for their Hezb-i-Islami connections. Dr Giustozzi states that the Afghan security services have a wide network of informers in the villages, passing on information about the movement of suspect individuals. He also expresses the view that the fact that the appellant’s father was with Hezb-i-Islami in the past would be seen by the security forces as a member of the pro-insurgency inclinators of the whole family.”

109. Yet, having set out this evidence, the judge’s immediate response in the next paragraph, [89], is to state “against this background, I have substantial doubts about the credibility of the appellant’s account that the security forces were looking for him”. If by this background he meant to include Dr Giustozzi’s evidence, then his statement that he has “substantial doubts about the credibility of the appellant’s account” is a non-sequitur . That background supported the appellant’s account. If the judge meant something different by the reference here to “background” he should have said so. In any event, the judge’s summary of Dr Giustozzi’s evidence leaves out the fact that the latter had also stated that in his opinion the photograph showed Hekmatyr with his “key men”. Since the judge went on to find that at best the photo showed the appellant’s father to be a mere Hezb-i-Islami member, it can be deduced that he disagreed with the expert, but no reasons are given. Presumably as well (in finding that the appellant’s whereabouts would have been well-known to the authorities for some time), the judge disagreed with Dr Giustozzi that it was plausible that the authorities only started paying an interest in the appellant when the situation in Nangarhar deteriorated in 2006-2007. Dr Giustozzi also considered there was a plausible explanation for the authorities only having made a “perfunctory” search of the appellant’s home on two previous occasions. Again the judge does not address that: he simply reached the opposite conclusion.

110. Whilst it is unnecessary to address the other grounds, we would note that we think that the judge did sufficiently engage with, and evaluate, the evidence of the other witnesses. We think the judge’s conclusions on the family tracing issue were viable. We see nothing in the contention that somehow the judge was bound by Singh v Belgium app.no.33210/11 to approach the issue of the reliability of the documents differently: see above paragraph 78.

111. For the above reasons the FtT judge erred in law and in ways making it necessary to set his decision aside.

112. We raised with the parties how, if we decided to set aside the FtT decision, we should proceed to deal with it. Mr Bazini submitted that in fairness we should remit it to the First-tier. Such a course would, he said, be in accordance with the Senior President’s Practice Statement. We observe that the Practice Statement (which has now been amended with effect from 13 November 2014) does not prescribe which course of action is to be adopted and that paragraph 7.3 states: “ Re-making rather than remitting will nevertheless constitute the normal approach to determining appeals where an error of law is found, even if some further fact finding is necessary.” Whilst in the appellant’s case the nature of the judge’s errors are such that no finding of fact can be preserved, we are conscious that this case has now been in the appellate process for a considerable time and was in fact remitted to the First-tier once before, on 21 November 2012, an error of law having been found in an earlier determination by a FtT on 3 August 2011.

113. If the only consideration were speed, remittal might ordinarily result in a quicker rehearing, but we intend to give directions to ensure that the case is kept in the Upper Tribunal and is expedited.

114. Mr Bazini submits that to decline to remit would be to deny the appellant his full panoply of onward appeal rights, but that is true of any case retained in the Upper Tribunal when an error of law is found and the decision is set-aside.

115. Mr Bazini appeared to submit at one point that whichever course we chose to adopt we should preserve the findings made by Judge Knowles in relation to the witnesses which, he says, were not challenged. Leaving aside that the judge’s determination does record concerns about the evidence of the witnesses as well as the appellant (see e.g. [86], line 17), we consider that given our reasons for concluding that the judge erred in his overall approach to credibility regarding events which took place when the appellant was a minor (which the witnesses also sought to give evidence about), it would be quite wrong to try and pick and choose certain findings and say they were unaffected by this error.

116. For the above reasons:

The First-tier Tribunal erred in law in a way necessitating that we set his decision aside.

The case is to be listed before the Upper Tribunal with directions that this (re-) hearing be expedited.

Pursuant to Rule 14 of the Tribunal Procedure (Upper Tribunal) Rules 2008 (SI 2008/269) I make an anonymity order. Unless the Upper Tribunal or Court directs otherwise, no report of these proceedings shall directly or indirectly identify the Appellants. This direction applies to both the appellants and to the respondent and a failure to comply with this direction could lead to Contempt of Court proceedings.

Signed Date

Upper Tribunal Judge Storey