GS v THE SECRETARY OF STATE FOR THE HOME DEPARTMENT

What follows is the determination of the panel. Whilst bearing another signature, the principal work on it, comprising an almost complete draft, was undertaken by Senior Immigration Judge Mather who, since being taken ill, has confirmed that he agrees with the conclusion. The determination concerns the application and scope of Article 15(c) of the Qualification Directive (2004/83/EC) (“the Qualification Directive”), incorporated into the Immigration Rules HC395 in paragraph 339C. In Part 7 of the determination we have summarised our findings, and the guidance from both the European Court in Elgafaji v Staatssecretaris van Justitie (C-465/07) (reported at [2009] 2 CMLR 45), and the Court of Appeal in QD and AH v Secretary of State for the Home Department [2009] EWCA Civ 620.

We have considered the general conditions in Afghanistan. In Part 9, we have explained why, at present, there is not, as a general matter, a serious and individual threat to a civilian’s life or person by reason of indiscriminate violence in situations of international or internal armed conflict.

We deal with access to Jalalabad in Part 10, and internal flight to Kabul in Part 11. There is a discussion about enhanced risk categories in Part 12.

The appellant is a citizen of Afghanistan. He originally claimed to have been born on 1 January 1990, but, following a social services assessment, a date of birth of 17 August 1989 was recorded. The Immigration Judge who originally heard his appeal observed that, either way, the appellant was over 18 at the time of the appeal.

The appellant arrived in the United Kingdom unlawfully on 3 August 2005. He applied for asylum shortly thereafter but, on 28 November 2007, his application was refused. The formal immigration decision was to remove him as an illegal entrant, with an indication that directions would be given for his removal to Afghanistan.

The appellant appealed. His appeal was heard by Immigration Judge B Lloyd on 18 January 2008. In the determination which followed, the appeal was dismissed on refugee, human rights and humanitarian protection grounds.

The appellant successfully applied for reconsideration.

At an earlier reconsideration hearing, before Senior Immigration Judge Mather sitting alone, both parties agreed that the Immigration Judge had made a material error of law. That is, that when he considered the question of humanitarian protection he did not deal with Article 15(c) of the Qualification Directive. In other words he did not consider` whether the appellant would be at real risk of serious harm as the result of “a serious and individual threat to a civilian’s life or person by reason of indiscriminate violence in situations of international or internal armed conflict”.

Following that hearing, the Tribunal issued directions, which were also published as country guidance. The directions recorded the fact that, in a letter dated 7 January 2009, the respondent conceded that:

“for the purposes of International Humanitarian Law (‘IHL’), there is at present an internal armed conflict in Afghanistan, and that for the purposes of IHL, the whole of the territory of Afghanistan is to be treated as being in such conflict.”

For reasons which we will give later, that direction is now of little relevance and consequently no longer falls to be treated as country guidance.

When the matter came before us, Mr Symes did not seek to suggest that the passage of time since the Immigration Judge’s determination was such that we should reopen the question of whether the appellant is entitled to be recognised as a refugee, because of any change of circumstances. The matter proceeded before us only on humanitarian protection grounds, and in particular the provisions of Article 15(c). The appellant did not give evidence at the reconsideration hearings. Oral evidence was given by Professor Farrell (see Part 8 below). Although we refer to Article 15(c) throughout much of this determination, because we deal with the European Court’s judgement in Elgafaji, in the United Kingdom Immigration Judges are concerned with the similarly worded provision in paragraph 339C, which incorporates Article 15(c) into our domestic law.

The Immigration Judge found that the appellant was not a credible or reliable witness.

The appellant had claimed that his parents had been murdered in furtherance of a land dispute with the local commander, Gul Karim, and the warlord, Hazrat Ali. He expressed fear that he would also become their victim. That was rejected. It was also the appellant’s case that his sister is married and living with her husband in Pakistan; and his brother died whilst in Dubai. The respondent accepts, as expressly conceded by Mr Palmer, that the appellant comes from Jalalabad, in the province of Nangarhar, and that he has no family remaining in Afghanistan. It is our task to consider this appeal on the basis that the appellant is a young man, about whom nothing has been proved save that he is from Jalalabad, is over 18, and has no immediate family in Afghanistan.

Mr Symes gave a very brief account of the recent history of Afghanistan, which we quote in order to put this determination into context. In paragraph 37 of his skeleton he said that:

“The conflict is protracted: Afghanistan has endured ‘almost constant warfare since the Soviet invasion in 1979’. The fracture of the Mujahedin following their successful overthrow of the puppet Soviet regime in 1992 led to civil war. After the formation of the Taliban in 1994, there was civil war between them and the Northern Alliance between 1996 and 2001. The current conflict was triggered by the US led invasion of Afghanistan in late 2001. The current phase of internal armed conflict has been ongoing since 2002.”

Paragraph 339C of the Immigration Rules HC 395 provides:

“339C. A person will be granted humanitarian protection in the United Kingdom if the Secretary of State is satisfied that:

(i)

he is in the United Kingdom or has arrived at a port of entry in the United Kingdom;

(ii)

he does not qualify as a refugee as defined in regulation 2 of The Refugee or Person in Need of International Protection (Qualification) Regulations 2006;

(iii)

substantial grounds have been shown for believing that the person concerned, if he returned to the country of return, would face a real risk of suffering serious harm and is unable, or, owing to such risk, unwilling to avail himself of the protection of that country; and

(iv)

he is not excluded from a grant of humanitarian protection.

Serious harm consists of:

(i)

the death penalty or execution;

(ii)

unlawful killing;

(iii)

torture or inhuman or degrading treatment or punishment of a person in the country of return; or

(iv)

serious and individual threat to a civilian's life or person by reason of indiscriminate violence in situations of international or internal armed conflict.”

The appellant is not excluded from a grant of humanitarian protection.

Paragraph 339C is in part derived from the Qualification Directive, including Article 2 and the definition of serious harm in Article 15(c).

The following provisions of the Qualification Directive are relevant, or referred to later:-

“Recital

(10)

This Directive respects the fundamental rights and observes the principles recognised in particular by the Charter of Fundamental Rights of the European Union. In particular this Directive seeks to ensure full respect for human dignity and the right to asylum of applicants for asylum and their accompanying family members.

(26)

Risks to which a population of a country or a section of the population is generally exposed do normally not create in themselves an individual threat which would qualify as serious harm.

Article 2

Definitions

For the purposes of this Directive:

…

(e)

‘person eligible for subsidiary protection’ means a third country national or a stateless person who does not qualify as a refugee but in respect of whom substantial grounds have been shown for believing that the person concerned, if returned to his or her country of origin, or in the case of a stateless person, to his or her country of former habitual residence, would face a real risk of suffering serious harm as defined in Article 15, and to whom Article 17(1) and (2) do not apply, and is unable, or owing to such risk, unwilling to avail himself or herself of the protection of that country;

Article 8

1.

As part of the application for internal protection, Member States may determine that an applicant is not in need of international protection if in a part of the country of origin there is no well-founded fear of being persecuted or no real risk of suffering serious harm and the applicant can reasonably be expected to stay in that part of the country.

2.

In examining whether a part of the country of origin is in accordance with paragraph 1, Member States shall at the time of taking the decision on the application have regard to the general circumstances prevailing in that part of the country and to the personal circumstances of the applicant.

3.

Paragraph 1 may apply notwithstanding technical obstacles to return to the country of origin.

Article 9

(3)

In accordance with Article 2(c), there must be a connection between the reasons mentioned in Article 10 and the acts of persecution as qualified in paragraph 1.

Article 15

Serious harm

Serious harm consists of:

(a)

death penalty or execution; or

(b)

torture or inhuman or degrading treatment or punishment of an applicant in the country or origin; or

(c)

serious and individual threat to a civilian’s life or person by reason of indiscriminate violence in situations of international or internal armed conflict.”

Article 17

Exclusion

1.

A third country national or a stateless person is excluded from being eligible for subsidiary protection where there are serious reasons for considering that:

(a)

he or she has committed a crime against peace, a war crime, or a crime against humanity, as defined in the international instruments drawn up to make provision in respect of such crimes;

(b)

he or she has committed a serious crime;

(c)

he or she has been guilty of acts contrary to the purposes and principles of the United Nations as set out in the Preamble and Articles 1 and 2 of the Charter of the United Nations;

(d)

he or she constitutes a danger to the community or to the security of the Member State in which he or she is present.

2.

Paragraph 1 applies to persons who instigate or otherwise participate in the commission of the crimes or acts mentioned therein.

The interpretation of Article 15(c) has not proved to be straightforward. The matter was referred to the European Court in Elgafaji. The Advocate General, in his opinion, said:

“Interpretation is not easy and is, to a large extent, open to debate…”

Having considered his opinion, and the arguments, the Court concluded first, that Article 15(c) is different from Article 3 of the ECHR which is essentially replicated in Article 15(b). Therefore, the Court said (at paragraph 28), Article 15(c) of the Directive must be considered independently of Article 3 of the ECHR.

In giving guidance on the interpretation of Article 15(c), the Court started by noting that the terms “death penalty”, “execution” and “torture or inhuman or degrading treatment or punishment of an applicant in the country of origin”, used in Article 15(a) and (b) of the Directive, cover situations in which an applicant is specifically exposed to a risk of a particular harm. It went on to say:

“33.

By contrast, the harm defined in art.15(c) of the Directive as consisting of a 'serious and individual threat to [the applicant's] life or person' covers a more general risk of harm.

34.

Reference is made, more generally, to a 'threat ... to a civilian's life or person' rather than to specific acts of violence. Furthermore, that threat is inherent in a general situation of 'international or internal armed conflict'. Lastly, the violence in question which gives rise to that threat is described as 'indiscriminate', a term which implies that it may extend to people irrespective of their personal circumstances.

35.

In that context, the word 'individual' must be understood as covering harm to civilians irrespective of their identity, where the degree of indiscriminate violence characterising the armed conflict taking place … reaches such a high level that substantial grounds are shown for believing that a civilian, returned to the relevant country or, as the case may be, to the relevant region, would, solely on account of his presence on the territory of that country or region, face a real risk of being subject to the serious threat referred to in art.15(c) of the Directive.

36.

That interpretation, which is likely to ensure that art.15(c) of the Directive has its own field of application, is not invalidated by the wording of recital 26 in the preamble to the Directive, according to which:

'risks to which a population of a country or a section of the population is generally exposed do normally not create in themselves an individual threat which would qualify as serious harm.'

37.

While that recital implies that the objective finding alone of a risk linked to the general situation in a country is not, as a rule, sufficient to establish that the conditions set out in art.15(c) of the Directive have been met in respect of a specific person, its wording nevertheless allows - by the use of the word 'normally' - for the possibility of an exceptional situation which would be characterised by such a high degree of risk that substantial grounds would be shown for believing that that person would be subject individually to the risk in question.

38.

The exceptional nature of that situation is also confirmed by the fact that the relevant protection is subsidiary, and by the broad logic of art.15 of the Directive, as the harm defined in paras (a) and (b) of that article requires a clear degree of individualisation. While it is admittedly true that collective factors play a significant role in the application of art.15(c) of the Directive, in that the person concerned belongs, like other people, to a circle of potential victims of indiscriminate violence in situations of international or internal armed conflict, it is nevertheless the case that that provision must be subject to a coherent interpretation in relation to the other two situations referred to in art.15 of the Directive and must, therefore, be interpreted by close reference to that individualisation.

39.

In that regard, the more the applicant is able to show that he is specifically affected by reason of factors particular to his personal circumstances, the lower the level of indiscriminate violence required for him to be eligible for subsidiary protection.”

Prior to the European Court’s analysis of Article 15(c), the Tribunal had considered its meaning on two occasions, in HH and Others (Mogadishu: internal armed conflict: risk) Somalia CG [2008] UKAIT 00022, and subsequently in KH (Article 15(c) Qualification Directive) Iraq CG [2008] UKAIT 00023. Those cases considered the meaning to be put upon various terms found in Article 15(c), and its relationship to Article 15(a) and (b). In so doing they concluded that the key terms should be given a meaning consistent with International Humanitarian Law (IHL). These cases were followed by AM & AM (armed conflict: risk categories) Somalia CG [2008] UKAIT 00091, which also considered the Advocate General’s opinion in Elgafaji.

Subsequent to the European Court’s judgment in Elgafaji, the Court of Appeal considered the interpretation of Article 15(c), in QD and AH. The Court disagreed with the notion that Article 15(c) should be interpreted with reference to IHL. In so doing they said that in KH (Iraq) the Tribunal’s approach had:

“led them to construe ‘indiscriminate violence’ and ‘life or person’ too narrowly, to construe ‘individual’ too broadly, and to set the threshold of risk too high.”

In paragraph 19, the Court identified three difficulties in interpreting Article 15(c):-

“(i)

the ostensibly cumulative but logically intractable test of ‘a real risk’ of a ‘threat’;

(ii)

the contradictory postulation of ‘individual threat’ to life or safety from ‘indiscriminate violence’;

(iii)

the requirement of ‘armed conflict’ when there may well be only one source of indiscriminate violence.”

Dealing with the interpretation of “a real risk of a threat” the Court said that, in the context of Article 15, a threat to a civilian’s life or person is concerned not with fear alone but with the possibility that it may become a reality. In paragraph 29, the Court said that:

“’Risk’ in Article 2(e) overlaps with ‘threat’ in Article 15(c), so that the latter reiterates but does not qualify or dilate the former.”

The Court also concluded that the word ‘serious’ in Article 15(c) is concerned with ‘threats of real harm’ and that the provision should be read as referring to ‘serious threats of real harm’. Quoting paragraph 136 of KH (Iraq), the Court also concluded (in paragraph 33), in relation to the degree of risk to individuals required to bring an armed conflict situation within the purview of Article 15(c), that all an applicant has to show is that incidents of indiscriminate violence:

“were happening on a wide scale and in such a way as to be of sufficient severity to pose a real risk of serious harm… to civilians generally.”

In considering the degree of risk required, the Court expressly said that it was not appropriate simply to read across a test which had been expressed in AA (Zimbabwe) [2007] EWCA Civ 149, namely a “consistent pattern of mistreatment”. The Court highlighted the contrast between methodical victimisation of those suspected of disloyalty, and the occurrence of indiscriminate violence. It said the risk of random injury or death, which indiscriminate violence carries, is the converse of consistency. In so doing it was making the point that the test is wider, and in so doing, approved that expressed in paragraph 136 of KH (Iraq).

Turning to what it had described as the contradictory postulation of individual threat arising from indiscriminate violence, the Court said that the effect of that phrase had been settled by Elgafaji.

Dealing with the third source of difficulty, relating to the requirement for armed conflict when there may well be only one source of indiscriminate violence, the Court, in paragraph 38, said that:

“the phrase ‘situations of international or internal armed conflict’ in Article 15(c) has an autonomous meaning which is broad enough to capture any situation of indiscriminate violence, whether caused by one or more armed factions or by a state, and which reaches the level described by the European Court in Elgafaji.”

The Court agreed with the UNHCR’s submission that there is no requirement for the armed conflict to be exceptional. The Court emphasised that the word ‘civilian’ means a genuine non-combatant, and not simply someone who is not in uniform. The Court observed that the UNHCR had submitted that former combatants should not be excluded from protection, but did not itself express a view. It was not a matter argued before us but we would consider that, whether a former combatant falls within the definition of a civilian, will be a question of fact and degree in every case in which the issue arises.

In conclusion, in paragraph 40, the Court defined the question which the Tribunal has to answer (adapted by us for this case) as being:

“Is there in [Afghanistan] or a material part of it such a high level of indiscriminate violence that substantial grounds exist for believing that an applicant such as [GS] would, solely by being present there face a real risk which threatens his life or person.”

By ‘material part’ the Court said it meant the applicant’s home area, or any potential place of internal relocation, if relevant. In our judgement, by so doing, the Court confirmed that it is possible, for the purposes of the Qualification Directive, to have an armed conflict in one part of a country, when other parts may be free of it. Similarly, even in an area of internal armed conflict, there may be parts where the high levels of indiscriminate violence, needed to obtain protection, are not achieved.

This is the first occasion since QD and AH in which Article 15(c) has come to be considered in detail by a panel of the Tribunal. We shall consider and apply the law as set out in Elgafaji and QD and AH, and address a number of issues which those cases did not purport to settle.

As a result of QD and AH, whether there is an internal armed conflict in Afghanistan for IHL purposes, is not a relevant consideration. For that reason, GS (Existence of internal armed conflict) Afghanistan CG [2009] UKAIT 00010 is no longer to be treated as extant country guidance.

The issues which remain to be considered, are:

In his skeleton, Mr Symes said little on these topics save to refer to the overriding purpose of Article 15(c), which he said was to give temporary refuge to people whose safety is placed in serious jeopardy by indiscriminate violence. He referred to paragraphs 23 to 34 of QD and AH, and particularly to paragraph 35 of Elgafaji, which describes the necessary degree of indiscriminate violence as reaching:

“such a high level that substantial grounds are shown for believing that a civilian, returned to the relevant country or, as the case may be, to the relevant region, would solely on account of his presence on the territory of that country or region, face a real risk of being subject to the serious threat referred to in Article 15(c).”

He expressed the view that the reasoning in Elgafaji may have been informed by the Temporary Protection Directive (Council Directive 2001/55/EC), as suggested in the UNHCR submission at paragraph 37.3 to the Appendix in QD and AH. There it was said that Directive:

“…applies to persons who have fled ‘armed conflict or endemic violence’ and persons at serious risk of ‘systemic and generalised violations’ of their human rights.”

In argument, Mr Symes considered the possible forms of threat to civilians. He referred to the requirement for a threat to life or person. In QD and AH, whilst the Court said that the IHL approach in KH (Iraq) had led to too narrow a construct of “life or person”, it did not venture an opinion as to what “threat to life or person” does mean. Mr Symes argued that Articles 15(a), (b) and (c) should be read together. He said that, whilst QD and AH did not mandate the suggestion that they should be considered euisdem generis,it did imply that was the case. He asked, rhetorically, whether casualties arising from disease or malnourishment would be included, and argued that is not ruled out under the provisions of Article 3 ECHR, which are replicated in paragraph 15(b). He wondered whether Sedley LJ had this in mind when he said, at the end of the judgment, that it was possible that QD may succeed in his claim on the basis of Article 15(c).

When the Tribunal reminded Mr Symes that there needed to be a causal nexus between the threat to a civilian’s life or person and the indiscriminate violence, Mr Symes suggested that where, for example, humanitarian aid supplies were interrupted by insurgents the consequence of that interruption may be sufficient. But, he conceded that there is a risk that those who were simply poor would not be successful.

Dealing with the question of causation, he reminded us that the wording in Article 15(c) is “by reason of”, whereas in the Refugee Convention causation is dealt with by the words “for reasons of”. He argued that as the language is almost exactly the same it is not possible to say that refugee cases are not relevant to issues of causation, when concerning subsidiary protection. He referred to Article 9(3) of the Qualification Directive which, in the context of the qualification as a refugee, provides:

“…there must be a connection between the reasons mentioned in Article 10 of the acts of persecution as qualified in paragraph 1.”

He argued that Article 9(3) did not use the word “exclusively”, or even “probably”, it simply required “a connection” and, he argued, a low level of connection. He then asked whether Article 9(3) would apply to subsidiary protection and suggested that the possibility should not be excluded. As IHL considerations have been ruled out, he suggested that an approach similar to that in Article 9(3) would be a good way of resolving what is required by way of causation.

He argued that the Qualification Directive’s overall approach is purposive and referred to recital 10 which says that the Directive:-

“seeks to ensure full respect for human dignity and the right to asylum of applicants for asylum and their accompanying family members.”

He acknowledged that this recital also refers to refugee claims but, he argued, the Directive is such that the recital covers both refugees and subsidiary categories. He argued that it is necessary to take a broad approach to the harm from which Article 15 was designed to protect, because recital 10 includes the reference to human dignity. He argued that it would be contrary to human dignity not to provide protection to those whose needs arise from either direct or indirect violence.

He referred to the UNHCR view that serious (including indiscriminate) threats to life, physical integrity or freedom, resulting from generalised violence or events seriously disturbing public order, are valid reasons for international protection under its mandate. This was said, not only in their January 2008 Position Paper, but also in paragraphs 50 to 55 of the submissions which appear in the appendix to QD and AH. That, he argued, also showed the need for a broad approach.

He also relied on Article 61 of the Consolidated Version of the Treaty Establishing the European Community, which he argued mandates the progressive establishment of an area of freedom, security and justice, including “safeguarding the rights of nationals of third countries”. He argued that Article 15(c) should not offer protection any less extensive, as to the forms of harm contemplated, than that protected by Article 15(b).

Mr Symes relied on the Advocate General in Elgafaji, at his paragraph 23, where he said:

“Accordingly, although the case law of the Strasbourg Court is not a binding source of interpretation of Community fundamental rights, it constitutes nonetheless a starting point for determining the content and scope of those rights within the European Union. Taking that case law into account is, moreover, essential to ensure that the Union, founded on the principle of respect for human rights and fundamental freedoms, will contribute to extending the protection of those rights in the European area.”

Mr Symes argued that when the Court of Appeal said, in QD and AH, that the Tribunal had erred in law by construing “indiscriminate violence” “too narrowly”, that amounted to an endorsement of the UNHCR approach. That approach requires all forms of violence, emanating both directly from the conflict itself and from other sources, such as opportunistic criminals taking advantage of the breakdown of law and order, to be taken into account. It followed, in his submission, that all forms and consequences of violence in Afghanistan should be deemed relevant to Article 15(c). Mr Symes added that, when the Court said that indiscriminate violence was broader than the IHL meaning, it could have ruled out categories of indiscriminate violence, but did not do so. In HH and Others (Somalia) (paragraph 333) the Tribunal said:

“…It is clear that the indiscriminate violence can comprise violence perpetrated by combatants, which fails to distinguish between civilian and military targets. But…the indiscriminate violence does not have to be violence that emanates directly from the combatants themselves. If that had been intended, we think the drafters could and would have said so. The indiscriminate violence may, for example, be perpetrated by looters and other criminal elements, taking advantage of a breakdown in law and order to go on the rampage.”

In paragraph 96 of KH (Iraq), the Tribunal drew back from that, because of their IHL approach, and said, in effect, that criminal violence taking advantage of the lack of law and order would not be included. Mr Symes argued that it would be, and the IHL approach had been found to be wrong.

In their skeleton argument, Mr Palmer and Mr Blundell referred to the jurisprudential history of Article 15(c). They noted that in QD and AH, although the UNHCR had argued against excluding criminal violence from the purview of the phrase “indiscriminate violence”, the Court of Appeal had not expressly adopted that part of the UNHCR’s submissions, in contrast to others. They therefore argued that the position with regard to criminal acts remains as set out by the Tribunal in paragraph 96 of KH (Iraq). There, it was said that criminal acts are capable of being sufficiently punished within the framework of the domestic criminal law of a country, even if that did not necessarily happen in practice; and that there is nothing in the text, the Directive, or the preparatory documents to suggest that protection against criminal violence was intended. That view is supported by the recital 26 which provides that risks to which a population is generally exposed are not normally enough to entitle an applicant to the protection of the Qualification Directive. They argued that nothing in QD and AH suggests that such acts should now be caught by Article 15(c). They asserted that, because none of the organisations which provide information about civilian casualties of war have attempted to provide figures for victims of criminal acts, that suggests they also do not consider it part of the indiscriminate violence to which Article 15(c) is addressed.

Mr Palmer dealt with the findings in Elgafaji. When he said that any situation of indiscriminate violence has to reach the high level described by the European Court, the Tribunal suggested to him that, for persons who were at heightened risk, there is a relatively low ‘Elgafaji test’, and asked how this impacted on his submissions. He said that depended upon what “indiscriminate violence” means. He argued that it is not any violence, or even major violence, but must be indiscriminate. He said that the issue is whether a person is at risk “by reason of indiscriminate violence in situations of international or internal armed conflict”. He suggested that a more helpful approach is to consider that as one phrase and ask what it means. Even if it is not to be construed strictly with IHL guidance, the violence must be looked at in the context of armed conflict. He then asked what the threshold is, saying that is bound up with the existence of an armed conflict. He suggested that the threshold was not that different from the type of violence which the Tribunal identified in KH (Iraq). Although we must take care, because in KH (Iraq) the analysis was in relation to IHL, he argued that that did not mean that all the analysis went out of the window and noted that, at paragraph 94, it was said that indiscriminate violence could still cover violence targeted against civilians directly, so that targeted violence is part of the general risk. Although QD and AH had found that “indiscriminate violence” had been construed too narrowly in KH (Iraq), he said that was not an obvious answer to the question. In QD and AH, the Court seemed to suggest that indiscriminate violence embraces armed anarchy, but he acknowledged that may well not display the nexus that was considered essential in KH (Iraq), although he did suggest it could explain the Delphic comments in QD and AH.

Mr Palmer agreed with the view in KH (Iraq), that criminal violence is not covered by “indiscriminate violence”. He said first, criminal violence is not indiscriminate and second, as had been accepted in HH & others (Mogadishu: armed conflict: risk) Somalia CG [2008] UKAIT 00022 there has to be a nexus to the armed conflict. To be included any criminal violence would have to be either caused, or permitted, by the armed conflict, for example, if there were a total breakdown of law and order. He suggested that any criminal activity had to be born of, and not simply not prevented by, the internal armed conflict. It was therefore something close to what the Tribunal had identified in KH (Iraq) and not tied to military operations. We note that QD and AH, at paragraph 21, included armed anarchy when speaking of a war zone.

Mr Palmer said that Mr Symes had read too much into the Court of Appeal’s remittal of the appellants’ appeals in QD and AH. He argued that, as the Tribunal is the fact-finding body, and as it had applied the wrong legal test, it was not for the Court of Appeal to apply the appropriate test to the facts as found, if it was not clear whether the evidence could meet the appropriate test or not. Not all the evidence about Iraq was before the Court in QD and AH.

When considering the degree of risk which exists, Mr Palmer discouraged us from applying a formulaic approach, such as relating a real risk to a one in ten chance (Batayav v Secretary of State for the Home Department [2003] EWCA Civ 1489). He argued that, in general, formulaic tests do not find favour, and the test remains that of real risk. He argued that it was completely inappropriate, as Mr Symes had done, to divide the population by the number of incidents. Even if that had been appropriate, it would not be done by taking Afghanistan as a whole. Whatever the test, the risk needs to be looked at area by area and case by case.

He suggested that Mr Symes had effectively said that the Tribunal should consider all the effects of war, because the casualty figures among civilians are so low, and did not reach anything like the one in ten criterion. If he did, Mr Palmer argued this was a wholly illegitimate approach, unsupported by anything in either QD and AH or Elgafaji.

Dealing with some of Mr Symes’ other submissions, he argued that “by reason of” simply meant there had to be a causal connection. He suggested that the approach in AM and AM was an appropriate one. At paragraph 93 of AM and AM, the Tribunal said the test was whether acts complained of were:

“one, albeit not necessarily the only, operative reason for the feared persecution.”

In AM and AM, Counsel had suggested an example of indiscriminate shelling of a civilian neighbourhood which caused death and wounding to civilians, arguing that those deaths and woundings would plainly be “by reason of the indiscriminate violence”. He submitted that, if in consequence of that violence, a surviving population was displaced to a region in which it was likely to die of starvation and disease, those consequences would also be “by reason of” the indiscriminate violence. The Tribunal accepted that, with a caveat. They said:

“In order for the indiscriminate violence to be an ‘effective cause’, it clearly cannot extend to include consequences that are connected only remotely.”

Mr Palmer submitted that the violence did not need to be the sole, but did need to be an operative, cause; and not too remote. He referred to Mr Symes’ submission that:

“Plainly there is ‘a connection between’ threats to life and person which are indirect effects of armed conflict and the indiscriminate violence which animates that conflict.”

and was anxious to say that, from that formulation, it does not follow that there would always be a causal connection. He gave as an example, a war which has caused food supply problems and said those problems are not an effect of indiscriminate violence. He emphasised that the test remains “by reason of indiscriminate violence”.

Dealing with Mr Symes’ other submissions, Mr Palmer argued that recital 10 of the preamble to the Directive is too vague to import the reading which Mr Symes asked us to adopt. It does not mean that Article 15(c) is there to protect human dignity. He reminded us of recital 26, which says that risks to which a population of a country or a section of the population is generally exposed do not normally create in themselves an individual threat which would qualify as serious harm. He said that, on Mr Symes’ view of the significance of recital 10, such risks would create individual threats.

As to Mr Symes’ assertions about the UNHCR view, he emphasised that the UNHCR is not a source of law when it came to the interpretation of the Qualification Directive.

Returning to Mr Symes’ assertion that, at the very least, Article 15(c) should offer protection no less extensive than the forms of harm protected by Article 15(a) and (b), Mr Palmer said that he did not accept that as a broad proposition. He said that because, in Elgafaji, the European Court said that Article 15(c) was different, and involved general harm, whereas 15(a) and 15(b) dealt with individual risk. He argued that Article 15(c), whilst protecting from more general harm, still has the individual component to it. As to the submission that the Advocate General in Elgafaji had said that the European Union would “contribute to extending the protection of [human] rights in the European Area”, Mr Palmer argued that this had been covered in QD and AH. It did not mean that there was to be a Europe wide protection from all the effects of war. In particular in Afghanistan, not all problems can be included. Much has been born of 30 years of armed conflict of differing degrees. For example, problems could be caused by such things as bridges being blown up, and that would not suggest indiscriminate violence.

Dealing still with indiscriminate violence, Mr Palmer asked whether pro-government violence could be described as indiscriminate in the context of the care which was now being taken by ISAF (to which we will come later). He said that even when steps are taken to avoid casualties among civilians, they will continue to occur. When they do, should the victims be properly described as victims of indiscriminate violence? He argued that he was not trying to import a quasi-IHL test, but said that the government-supported forces are using a high degree of discrimination. He suggested that, whilst the anti-government forces are far more indiscriminate, and a lot of anti-government activity can properly be described as indiscriminate, it is still necessary to determine the extent of the indiscriminate violence. He argued that evaluation is a matter of judgment, but suggested that the evidence is not sufficient, in most cases, to show that a person is the victim of indiscriminate violence. In particular, being caught in crossfire does not in itself demonstrate that one is the victim of indiscriminate violence.

During the course of discussing elevated risk categories, Mr Palmer was anxious to suggest that it was not a question of people being at risk because they are, say, teachers. He said that the common factor among high risk categories is that they are perceived collaborators, in one form or another. There has been evidence of warnings about teachers taking mixed, or female, classes but that is not the same as saying that all teachers are at risk on return. He emphasised that, if somebody is at risk because they are a member of a group, such as teachers, or even perceived collaborators, they may well have a refugee claim. Therefore, when considering Article 15(c) risk, care needs to be taken to ensure that, if appropriate, they should be recognised as refugees, rather than granted humanitarian protection.

Using the example that Mr Palmer gave, namely teachers who may be warned against teaching mixed classes, or classes of girls, we asked whether the approach taken in HJ (homosexuality: reasonably tolerating living discreetly) Iran [2008] UKAIT 00044 applies in this context. There the Tribunal (following a remittal from the Court of Appeal in J v SSHD [2006] EWCA Civ 1238) considered the extent to which it is reasonable to expect a person to avoid difficulties by modifying their behaviour upon return to their home country. Mr Palmer argued that the principles were no different. He said Article 15(b) effectively introduced Article 3 ECHR into the ambit of humanitarian protection and it was to Article 3 that the principle was applied. He saw no reason for it not to apply equally to Article 15(c). Mr Symes did not argue to the contrary.

We know, from the formulation in QD and AH, that the question to be answered by the Tribunal is that posed at paragraph 40:

“Is there in [Afghanistan] or a material part of it such a high level of indiscriminate violence that substantial grounds exist for believing that an applicant such as [GS] would, solely by being present there, face a real risk that threatens his life or person?”

Following QD and AH we know that, for the purposes of interpreting Article 15(c), armed conflict is not defined by reference to International Humanitarian Law and that it has:-

“an autonomous meaning broad enough to capture any situation of indiscriminate violence, whether caused by one or more armed factions or by a state, which reaches the level described by the European Court of Justice in Elgafaji” (paragraph 35 of QD and AH).

What this means is that the real focus of attention is on the intensity of the indiscriminate violence, rather than on the nature of the conflict giving rise to the situation in which such violence exists. Accordingly, the formulation of the question by Sedley LJ in paragraph 40 of QD and AH does not contain a reference to armed conflict. In QD and AH it was also accepted that there is no requirement for the armed conflict to be exceptional, although a degree of intensity of indiscriminate violence is required in order to pass the test in Elgafaji.

That test for the required degree of intensity appears at paragraph 35 of Elgafaji. Having referred to the assessment having to be made by the authorities or courts of a member state, a person is to be entitled to humanitarian protection when:

“The degree of indiscriminate violence characterising the armed conflict taking place…reaches such a high level that substantial grounds are shown for believing that a civilian, returned to the relevant country or, as the case may be, to the relevant region, would solely on account of his presence on the territory of that country or region, face a real risk of being subject to the serious threat referred to in Article 15(c) of the Directive.”

In paragraph 37, the court referred to “an exceptional situation which would be characterised by such a high degree of risk…”; although we observe the warning in QD and AH not to treat exceptionality as a legal requirement.

That test was finessed in paragraph 39 of Elgafaji, where the Court said:

“In that regard, the more the applicant is able to show that he is specifically affected by reason of factors particular to his personal circumstances, the lower the level of indiscriminate violence required for him to be eligible for subsidiary protection.”

In relation to that last proposition, the Court added that regard should be had to the area in which indiscriminate violence was occurring, the actual destination of the applicant, and whether Article 4(4) of the Qualification Directive applied. Article 4(4) provides:-

“The fact that an applicant has already been subject to persecution or serious harm or to direct threats of such persecution or such harm, is a serious indication of the applicant’s well-founded fear of persecution or real risk of suffering serious harm, unless there are good reasons to consider that such persecution or serious harm will not be repeated.”

The high degree of risk required to bring a situation within Article 15(c) is emphasised by the Court throughout paragraphs 36 to 38. The Court recognised that Recital 26 of the Qualification Directive provides that risks to which a population (or part) is exposed do not normally give rise to individual threat which would qualify as serious harm, but said that the word ‘normally’ allows for an exceptional situation characterised by such a high degree of risk that substantial grounds would be shown for believing that a person would be subject individually to that risk in question. That high level of risk, the Court said, is confirmed by the subsidiary nature of the protection. This was referred to in QD and AH where, in paragraph 37, the Court spoke of events occurring on a wide scale and being of sufficient severity to pose a real risk of serious harm.

QD and AH explained that Article 2 and Article 15(c) read together are concerned with serious threats of real harm.

An issue which has not been settled by the jurisprudence is the distinction between discriminate and indiscriminate violence. We do know, from QD and AH that indiscriminate violence does not include methodical victimisation. In paragraph 34 of Elgafaji, the Court said the term implies that the violence may extend to people irrespective of their personal circumstances. This appears to us to be a issue of fact in each case. Neither party attempted to define the difference any more precisely, but rather proceeded by way of example. It would, in our judgment, be indiscriminate violence if a suicide bomber were to attempt to assassinate one individual in a crowded market place. Similarly, the bombing of insurgents who were sheltering in a school, or other area known to be populated by civilians, would be indiscriminate. On the other hand, a targeted attack on opposition fighters, which unexpectedly caught individuals in the crossfire would not. More problematic areas may include the use of improvised explosive devices intended to attack coalition forces, but where explosions may occur in such a way that civilians are affected. Another problematic example is the incident we refer to later where a man driving a car bomb, intended to be used against a coalition convoy, unexpectedly collided with a tractor and detonated the bomb, thereby killing civilians nearby. The extent to which driving a car, that is in effect a bomb, through a crowded place on the way to attack coalition forces is an act that is so reckless that any adverse consequences are indiscriminate, was not debated before us.

The question of whether criminal activity falls within the scope of Article 15(c) is an issue which has been considered before. KH (Iraq) purported to largely rule it out (paragraph 96), but that was because its definition of armed conflict was based on IHL considerations, and we now know that was wrong. In HH and others (Somalia) it was recognised that indiscriminate violence does not have to be violence that emanates directly from combatants themselves. The Tribunal said:

“The indiscriminate violence may, for example, be perpetrated by looters and other criminal elements, taking advantage of a breakdown in law and order to go on the rampage.”

Although he accepted that QD and AH could be read so as to include armed anarchy within the meaning of indiscriminate violence, Mr Palmer argued that such violence may fail the nexus test. He also argued that there is nothing in the text of the Directive, or in the preparatory documents to suggest that criminal violence was intended to be included; that risks to which the population are generally exposed are not normally enough (recital 26 of the Qualification Directive); and that criminal acts would be capable of being sufficiently punished within the framework of domestic criminal law of the country, even if this may not necessarily happen in practice. We do not agree with his submission that paragraph 96 of KH (Iraq) has survived, because the reasoning therein was related to the need to protect civilians from violations of IHL. That paragraph did consider, as an example, criminal gangs exploiting the law and order vacuum by stealing oil from a pipeline, during an armed conflict, and thereby killing innocent civilians indiscriminately. That is an interesting example because it illustrates the difficulty of deciding whether the indiscriminate violence has been caused by one or more factions, or a state. There have been examples of that activity causing considerable loss of life and injury to innocent people in a complete absence of armed conflict.

We see no reason in principle why criminal acts should not be included in the scope of indiscriminate violence and, indeed, it is often difficult to separate armed conflict from a criminal act. It is hard to envisage an act more criminally culpable than carrying and detonating a bomb in a crowded marketplace, whatever the intention of the person concerned. Similarly, could it properly be argued that the roadblocks set up for reasons of extortion around Mogadishu in recent years are not the consequence of a complete breakdown in law and order arising from the armed conflict which is manifestly occurring there? The correct approach is not simply to ask whether the indiscriminate violence is criminal, or in pursuance of the armed conflict. It is a question of causation. The words used in Article 15(c) are “by reason of indiscriminate violence in situations of international or internal armed conflict”. There therefore needs to be a causal link between the threat to life or person and the indiscriminate violence, but that indiscriminate violence does not need to be caused by one or more armed factions or the state. We emphasise that, criminal acts, as with any other form of indiscriminate violence, need be of sufficient severity to pass the Elgafaji test, and produce a serious and individual threat to a civilian’s life or person (to which we will come). Not all criminal acts, by a very long way, would fall into that category.

What is the level of nexus required between the serious and individual threat and the indiscriminate violence? We can do no better than to adopt the Tribunal’s test in AM and AM, where it was accepted that there is no significant distinction between the words “by reason of” and “for reason of”. The Tribunal said that to succeed an applicant must show that the indiscriminate violence is an effective cause of the serious and individual threat. It does not need to be the only cause, but has to be more closely connected than only remotely. That is also consistent with Article 9(3) of the Qualification Directive, which Mr Symes submitted should apply to the assessment of entitlement to humanitarian protection as it does to refugee status.

The next question is whether indirect consequences of indiscriminate violence can be sufficient to bring a person within Article 15(c). In our judgment this is also answered by considering the causal nexus, and will be a question of fact in each case.

We look at the issue in the context of Afghanistan. In doing so, we have regard to the evidence adduced in this case, which we describe in more detail later. In that country a significant proportion of the population does not have sufficient food and significant numbers have died of starvation. There has been an armed conflict for many years, although it has not been consistent in its severity, and has not always involved the same parties.

One consequence of the years of conflict is that agriculture, and food distribution, have suffered and that has given rise to difficulties of food supply. In our judgment it cannot be said that such a general situation has come about “by reason of indiscriminate violence in situations of international or internal armed conflict”. The food supply difficulties arise from a situation that has gone on for many years, and have not been shown to be the result of indiscriminate violence, as opposed to the targeted violence of armed groups against one another. Also, there is no satisfactory evidence that, even without an armed conflict, the situation in Afghanistan in this regard would be a great deal better. The food supply problem cannot be shown to be connected otherwise than very remotely to indiscriminate violence, even if it is more closely connected to armed conflict.

On the other hand, Mr Palmer’s example of indiscriminate bombing which leads to the population of a particular village having to flee to an area where they cannot be fed, could be said to have a causal nexus with the indiscriminate violence. The distinction is underlined by recital 26 of the Qualification Directive which provides that:-

“Risks to which the population of a country or a section of the population is generally exposed do normally not create in themselves an individual threat which would qualify as serious harm.”

That brings us to Mr Symes’ assertion, that such risks are included within the definition of Article 15(c) by virtue of recital 10 of the preamble to the Qualification Directive. That makes reference to seeking to ensure full respect for human dignity. We do not accept his arguments that recital 10, when read with Article 61 of the Consolidated Version of the EU Treaty, means that questions of human dignity have to be read into Article 15(c). We prefer Mr Palmer’s view that recital 10 is too vague to import that reading into Article 15(c). Recital 10 is an aspiration. Nor do we take the simplistic approach, that Article 15(c) should be seen to offer protection no less extensive, as to the forms of harm comprehended, and protection contemplated, than Article 15(b). We say that because, in Elgafaji, it was said that Article 15(c) is not the same as Article 15(a) and (b) in as much as 15(a) and (b) deals with individual risks, and Article 15(c) to a more general risk (despite its use of the word ‘individual’. An individual risk, and a general risk, are not the same and it is not therefore appropriate for the definition of one, to inform the definition of another. We find that the UNHCR view, whilst to be respected, is not persuasive of the interpretation of the Qualification Directive.

The next issue is the meaning of “life or person”. Clearly this must mean more than just a risk of death; if not, the words “or person” would have been unnecessary. In KH (Iraq) the phrase was subjected to an IHL analysis and that does not assist. In QD and AH the only relevant finding is that “life or person” was construed too narrowly by the Tribunal in KH (Iraq). Mr Symes, in his skeleton, suggested that a less narrow approach, and one which he suggested was impliedly found to be appropriate in QD and AH, is that seen in the UNHCR submission in that case. There, it was suggested that the forms of harm covered by Article 15(c) include all forms of serious physical and psychological harm, including flagrant breaches of qualified rights, such as freedom of thought, conscience and religion. We do not accept that the threat to a civilian’s life or person envisioned in Article 15(c) goes anywhere near as far as Mr Symes argues. As we have already said, the UNHCR is not an authority on the interpretation of the Qualification Directive. The forms of harm suggested by UNHCR are broadly consistent with those from which protection is required by the European Convention. They are more appropriately dealt with, if they reach ECHR Article 3 levels, or otherwise are sufficiently flagrant, by Article 15(b). They are also more in the nature of individual risks and unlikely to arise as a result of indiscriminate violence. Having said that, there is no precise definition that can be applied to the words “or person”. In our judgment the threat to a civilian’s person, which may bring him within the scope of Article 15(c), must be informed by the requirements in Articles 2(e) and 15(c). An applicant must show that there are substantial grounds for believing that he would face a real risk of suffering a serious and individual threat. It is clear from Elgafaji that there is a high threshold, and we know from QD and AH that Article 15(c) is concerned with “serious threats of real harm”. The meaning of “or person” must be read in the light of the need for “real harm” or “serious harm”. In that regard, it is possible to read across the provisions of Article 15(b) to understand the level (but not type) of harm required to obtain the protection of Article 15(c).

Another issue for a decision maker is the interrelationship between “individual threat” and “indiscriminate violence”. That was answered in paragraph 35 of Elgafaji (read with paragraph 39) and which we have already quoted. For the sake of completeness, we repeat it here:

“35.

…the word ‘individual' must be understood as covering harm to civilians irrespective of their identity, where the degree of indiscriminate violence characterising the armed conflict taking place…reaches such a high level that substantial grounds are shown for believing that a civilian, returned to the relevant country or, as the case may be, to the relevant region, would, solely on account of his presence on the territory of that country or region, face a real risk of being subject to the serious threat referred to in Article 15(c) of the Directive.

39.

In that regard, the more the applicant is able to show that he is specifically affected by reason of factors particular to his personal circumstances, the lower the level of indiscriminate violence required for him to be eligible for subsidiary protection.”

There is one further matter with which we should deal, and that is the person with a higher level of risk. In Elgafaji, the Court referred to recital 26 of the Directive, where it is said that risks to which a population, or a section of the population, of a country is generally exposed, do not normally create in themselves an individual threat which would qualify as serious harm. The word “normally” was said, in Elgafaji, to allow for the possibility of an exceptional situation characterised by such a high degree of risk that substantial grounds could be shown for believing that that person would be subject, individually, to the risk in question. From that the Court concluded in paragraph 39:

“…the more the applicant is able to show he is specifically affected by reason of factors particular to his personal circumstances, the lower level of indiscriminate violence required for him to be eligible for subsidiary protection.”

There is therefore a moving standard in the required level of indiscriminate violence, when a person is at a higher degree of risk. But, ultimately the test is one of real risk. That is clearly demonstrated by the wording of Article 2(e), which makes reference to “substantial grounds…for believing…would face a real risk”. The varying standards of violence and risk must be assessed within the overall criteria in Article 15(c).

In assessing the current situation in Afghanistan, we have had the benefit of a report, and oral evidence, from Professor Farrell; a report prepared in December 2008 by Dr Antonio Giustozzi; and a wealth of documentary evidence produced by each of the parties. We reminded the parties that, unless they referred expressly, either by way of skeleton arguments or submissions, to any particular piece of the documentary evidence they could not be certain that we would consider it, or assume that they relied upon it.

Dr Giustozzi did not give oral evidence to the Tribunal but his written report was before us. Dr Giustozzi has been accepted by the Tribunal as an expert on Afghanistan in a number of previous cases. He visits Afghanistan regularly. His report in this appeal was, however, of limited value because it was prepared on the assumption that the appellant’s expressed fear of Gul Karim and land grabs was correct.

With regard to general conditions, the two important parts of Dr Giustozzi’s report for present purposes are those which were put to Professor Farrell in cross-examination (paragraph 100 post). We shall describe those parts in the context of that cross-examination. Dr Giustozzi also dealt with some aspects of general conditions in Kabul, with which we will deal when we come to look at internal relocation.

Professor Theo Farrell is Professor of War in the Modern World, in the Department of War Studies, King’s College London. His CV shows that he is also Head of Military Studies Research for the King’s War Studies Group (which covers Defence Studies in the Department of the Joint Services Command and Staff College); a Consultant to the UK Ministry of Defence; and an Associate Fellow of the Royal United Services Institute. His report was designed to assist with three issues:

To a large extent the second issue is a matter for us, and is relevant to the extent that it is a factor in the definition of ‘indiscriminate violence’ in accordance with Elgafaji, as explained in QD and AH. Mr Palmer did not seek, for our purposes, to argue that Afghanistan is not in a state of internal armed conflict. The previous concession, that it is, for International Humanitarian Law purposes, is no longer relevant.

In his report, Professor Farrell gave a brief history of the academic study of war and an explanation of the development of the “human security paradigm” in security studies. He explained that the paradigm is a relatively new way of thinking about security, not in terms of securing states, but in terms of providing security for people. The definition of security for people includes freedom from starvation and oppression as well as violence. This has led to a change in approach. The assessment of the severity of a war used to involve estimates of the number of war dead, and traditionally the emphasis was on numbers of military personnel killed. The focus has now changed and has been directed to include civilian casualties, among other indicators of the adverse impact of war on human security.

Having explained how a war was defined for these purposes (more than 1,000 military personnel dead; no period was specified), he explained that the interest of academics has now moved to include intra state conflicts, as well as inter-state disputes.

He suggested that one way to measure the severity of a war is by the number of civilians killed and injured. It is especially appropriate for internal armed conflicts, which often do not involve battle between recognised military forces, but often do involve violence directed at civilian populations. He gave as examples three internal conflicts where the battle death threshold was less than 1,000 but where there were massive civilian losses. These were: 2 million dead in Cambodia between 1975 and 1978, 350,000 dead in Somalia between 1990 and 1991, and 800,000 dead in Rwanda in 1994. Professor Farrell explained that this “metric” can be problematic in application, because data on civilian casualties can be variable in both substance and reliability. He said this may be because parties to the conflict are less interested in gathering accurate data on civilian casualties, than in manipulating the data to support their own view of the conflict. The data provided by non-governmental organisations and other independent studies can also be very variable, with different researchers providing different casualty figures for the same conflict (he cited a difference in Iraqi casualties, one based on the Iraq Body Count Project, and the other on a study published in the Lancet). Another difficulty in estimating civilian casualties is that figures often do not distinguish clearly between combatants and civilian casualties, not least because in many contemporary conflicts the line between a combatant and a civilian is blurred. Also, estimates vary between direct and indirect casualties. He described direct casualties as those killed and injured in fighting, whereas indirect casualties included those who were killed or suffered serious injuries as a result of the effects of war. He gave examples of the latter, citing imprisonment, abuse, starvation or the destruction of critical infrastructure and services. He suggested that in order to fully appreciate the severity of a conflict, it is necessary to include indirect casualties. He illustrated that by saying that of the 2 million civilians that died in Cambodia, only 80,000 to 100,000 were directly killed. The rest died of starvation and disease, because of the policies of the Pol Pot regime. He summarised by suggesting that the civilian casualties provide a truer estimate of the severity of an armed conflict than battle casualties, whilst acknowledging that compilation of reliable data on civilian casualties involves challenges and methodological choices.

Professor Farrell also considered that two other consequences of conflict should be taken into account as possible metrics. The first is the number of displaced people, including both refugees and internally displaced persons (IDPs). As, in this jurisdiction, the term refugee is a term of art, and Professor Farrell confirmed in evidence that by refugee he meant a person who was displaced, but had crossed an international border, we propose to use the acronym “EDP” (externally displaced person) to describe such people. He noted that, on three occasions, United Nations Security Council Resolutions had been made when flows of EDPs became recognised as constituting a threat to international peace and security. In two cases international armed intervention took place despite a relatively low number of civilian deaths prior to intervention (in Haiti between 1991 and 1994, and in Kosovo). He considered that numbers of displaced people was a helpful measure, because reliable comparative data on EDPs and IDPs are available from the UNHCR. The other metric is state failure. He said that state failure, when chronic, can lead to the collapse of infrastructure and basic services, including law and order, making life for communities unsustainable. He recognised that state failure can be a causal factor in internal armed conflict, but emphasised that armed conflict can also be the cause of state failure. He warned about using mass population displacement, and state failure, as independent tests of conflict severity. He said mass population displacement and state failure can coexist, with mass population displacement being an indicator of state failure, and possibly a further contributor to it. He argued that, whilst these two measures may provide reinforcing evidence when looking at the severity of an armed conflict, they are not necessarily independent tests of conflict severity.

Professor Farrell therefore suggested that four metrics should be applied to determine the severity of the conflict in Afghanistan and the consequent risk to civilians, namely (i) battle deaths, (ii) civilian casualties, (iii) population displacement, (iv) state failure.

Turning, in his report, to an assessment of the severity of the conflict in Afghanistan, and the risk to civilians, Professor Farrell said that military battle deaths had remained relatively low. The figures he gave were 628 US military killed in action and 402 non-US NATO forces, between 2001 and 2008. He did not have any reliable figure for battle deaths from the Afghan National Army. News reports suggest that the Taliban have suffered far higher numbers, with an estimate from the Ministry of Defence suggesting 6,000 to 7,000 Taliban killed in action between 2006 and 2007. As to civilian casualties, the estimate is that 2,118 civilians were killed in fighting between pro-government forces and insurgent groups in 2008. Prior to that, 2006 had been regarded as the highest recorded year for civilian casualties since 2002. He said that those figures do not include civilians injured by the Taliban and Al-Qaeda in their campaign of violence and intimidation designed to terrify and subdue Afghans. That occurs especially in the South and East to prevent civilians from supporting the democratically elected government. According to the Afghan Independent Human Rights Commission, the campaign involves abduction, murder, mutilation and maiming of civilians and is especially directed at those accused of spying or collaboration, as well as local officials, police, teachers, medical staff and aid workers, with a terror campaign of suicide and other bomb attacks that is increasing in intensity. He said Human Rights Watch have estimated that up to 1,000 civilians had lost their lives as a result of Taliban and Al-Qaeda bombings in 2006, with an increasing trend towards improvised explosive devices (“IEDs”) and suicide bomb attacks in 2007/2008.

Tuning to population displacement, Professor Farrell said that throughout 2006 and 2007 Afghanistan was the leading country of origin for EDPs, producing 3.1 million or 20% of the global EDP population. This figure equates to about 10% of the population of Afghanistan and, although refuge was sought in 72 countries, 96% went to Pakistan or Iran, according to UNHCR. It is thought that, as of May 2008, there are 150,000 IDPs in Afghanistan, of which 110,000 come from the South of the country.

As to state failure, Professor Farrell gave a history of attempts to establish democratic central government. He said that the Brookings Institution in Washington DC, a highly respected think tank in Washington which produces a monthly report called the Afghan Index, ranks Afghanistan as second, only after Somalia, on its index of state weakness. Not all measures put it in the same position but, whilst attempts continue to improve the situation, the impact on sustainable living is evident. According to Brookings, only one in five Afghans has access to safe drinking water, and only 12% had access to adequate sanitation in 2008. 45% of Afghans experience food poverty, and infant mortality runs at 13% (20% for under 5s).

In drawing his conclusion, Professor Farrell said that by any measure Afghanistan is in severe conflict but that conflict is producing relatively few deaths among the NATO military. A conservative estimate puts the number of Taliban and Al-Qaeda deaths in the low thousands each year between 2006 and 2008. That number exceeds the battle deaths threshold, using the most widely cited statistics and most definitions of war. As to civilians, he said the number killed in the fighting is low in comparison with other conflicts, although the Taliban and Al-Qaeda campaign of terror and intimidation has killed and injured many more. With regards to population displacement, it is extremely high, with 10% of the population displaced externally, mostly to Pakistan and Iran. He considered that Afghanistan suffers from chronic underdevelopment of local government infrastructure and social services. Infant mortality and illness from disease and malnourishment, from thirty years of war, were adding indirect casualty figures which would be very high. He said that the situation appears to be getting worse. He quoted the UN Secretary General, reporting in September 2008, that his impression was that the situation in the country had deteriorated over the previous six months, with 40% to 50% of the country inaccessible to UN aid activities, and an upward trend in insurgent attacks against aid workers and convoys. He noted the intensification of insurgent influence, in areas that were previously calm, including provinces close to Kabul. He said the Brookings Institution recorded a 48% increase in attacks by Taliban and Al-Qaeda in Kabul in 2008; and a 51% increase across the whole of Afghanistan, but with attacks concentrated in the South and East of the country. Nangarhar Province suffered 292 terrorist and insurgent attacks in 2008, in comparison to 157 in Kabul. In concluding, he said the direct risk to civilians of violence from insurgents, terrorists and criminal gangs will remain very high, and the indirect risk to civilians, associated with chronic state failure, will remain extreme.

He appended to his report, at page 23, a comparison of incidents carried out by Taliban/anti-government elements (“AGEs”) by province in the first 39 weeks each of 2007 and 2008. The figures for Kabul were 106 and 157 respectively; for Nangarhar 244 and 292 and, as a comparison, Kandahar 534 and 820. Although percentage changes were provided we have not quoted those because when the figures are low, for example in Kabul, the percentage change can be misleading.

In evidence, Professor Farrell said that much of his statistical information came from the Brookings Institution. He was asked by Mr Symes about the background causes for the lack of sustainable living conditions. He said that it was 30 years of war, starting with the 79-89 war between the Mujahedin and the Soviet-backed regime, mostly in the Pashtun areas. He reiterated that it is now consequences of conflict which are looked at particularly carefully in security studies. Interest lies no longer just in those directly killed or involved in fighting, but rather in what is required for sustainable living, what causes damage to the infrastructure, and what may prevent sustainable living.

He was asked whether this concept of the human security paradigm had received “traction” outside the academic community, and said that it had. It has attracted the attention of the United Nations which, through Kofi Annan, has set up an organisation to develop the concept. The British Government position has been informed by the paradigm, and there has been a section in the Foreign and Commonwealth Office Annual Report on the responsibility of British forces to protect, which was clearly informed by the paradigm. He said that it also informs the Ministry of Defence who make reference to “the force for good”.

In cross-examination, he said military officers have been persuaded to think of war very differently. The first real lessons were learned in Bosnia, where troops were led by the UNHCR. The key lesson is that victory is not achieved by defeating the enemy, but by creating conditions for a sustainable peace. The new approach was therefore an “effects based approach”. He said that, as a formal approach it has in fact recently been abandoned by the United Kingdom, although the underlying philosophy has not. He said that the idea is to produce a security umbrella of both aid and law and order.

The practical effect has been that US operations increasingly emphasise the importance of avoiding civilian casualties, and that Commanders in ISAF have been told to exercise restraint in air raids. He said that, although the risk from air strikes is a very small element of risk in Afghanistan, the change can make a huge political difference.

Professor Farrell was asked about some of the statistics he relied upon. He agreed there was a complication in the estimated casualties as between combatants and civilian casualties. There were very small numbers of insurgent casualties caused by suicide attacks. Even though groups of insurgents intent on an attack do take shelter in civilian areas with their mortars, ISAF has a problem in assessing casualties, because the Taliban remove their dead and any bodies that are found are probably civilian. He was unable to say how reliable the figures are for civilians and combatants. He thought that the ISAF figures probably err on the side of caution, as they have no interest in exaggerating the figures. He thought the figures for Kabul would be more reliable than the overall figures, and said his qualification about the validity of the figures has more application in the South of the country.

Asked about his second qualification, that concerning direct and indirect casualties, he thought that the figures from The United Nations Assistance Mission to Afghanistan (UNAMA) probably referred to direct casualties and did not attempt to measure indirect casualties.

When asking about the risk to civilians by reason of indiscriminate violence, Mr Palmer suggested that there is a wider effect than the war itself. Professor Farrell accepted that a complicating aspect is the rise of warlords over the last twenty years. He said there is still a struggle between warlords and the central government, and a lot of violence is directed towards individual families. He argued that the warlords are part of the internal armed conflict but, as some warlords are also ministers, it is not something that is often recognised. He said that the causes of indiscriminate violence vary throughout the country: part is the NATO campaign, part is the intimidation and retribution by the Taliban and, elsewhere, warlords. He was asked whether the effects of war are simply that: the effects of war; and not of indiscriminate violence. He said that the level of problems is the consequence of the war since 1979. Prior to that Afghanistan was no worse than any other Central Asian country, now it is much worse. He said there have been different phases and that, since 2008, there has been an increase in the use of IEDs and terrorist attacks, and a decline in ambushes. He accepted that the recent increase in battle casualties was linked to an increase in offensive activities and the increase in ISAF troop numbers.

Professor Farrell was taken to a chart on page 17 of his report. That showed there had been 1,445 civilian deaths as a direct result of fighting between pro-government forces and AGEs in the first eight months of 2008. He accepted that figure was low, compared with other conflicts around the world. He said the figure did not include civilians deliberately killed and injured by the Taliban and Al-Qaeda. He was then taken to the figures produced by UNAMA in January 2009. These did deal with civilian casualties caused by AGE actions and showed 62% of the 1,160 victims had been as a result of suicide and IED attacks, and 23% from assassinations. 1,160 amounted to 55% of the total casualties for 2008. Of the AGE actions, the balance of casualties (15%) arose from rocket attacks and ground engagements in which civilian bystanders were affected. In other words, he accepted, the latter were caught in traditional crossfire, or because they were too close to a target. He accepted that the assassination figures represented targeted collaborators. Dealing with his acknowledgement that, in the light of all this, the figures for civilian deaths were low, he said this war does not “hit a lot of civilians”, but it is necessary to look at second order effects; in particular, the fact that 10% of the population has fled across the border. He accepted that according to the Brookings Institution, 4 to 5 million people had returned to Afghanistan since 2002, but said that EDP numbers were still very large, although the IDPs were a relatively small number.

Professor Farrell acknowledged that the data show the trend in increased Taliban and Al-Qaeda IED and suicide bomb attacks from 2007 to 2008. Figures were given for this in a table 1.3 from the Brookings Institution copied into his report at page 13. Professor Farrell agreed that he drew the inference of the increasing trend from the US army fatalities shown there.

The table on page 23 of his report, to which we have already referred, contains the comparison of incidents carried out by Taliban and anti-government elements, by province. He accepted those figures did not distinguish between attacks on troops and those on civilians. He said that one cannot analyse the levels of danger to civilians by province, otherwise than by extrapolating the national figures. He said the totals are increasing, and from that he inferred an increase in violence between the two sides.

Asked whether there were up-to-date figures, for incidents by province, Professor Farrell said there were up-to-date figures in the latest Brookings Institution report, but it now deals only with IEDs, and not all incidents. He said that report showed that in Nangarhar, in the first six months of 2009, there were 89 IED events, which was up 71%. He did not know how reliable that data is. He did however accept that an ‘IED event’ includes the discovery of an IED, and does not imply that there were any casualties. Professor Farrell said that he has recently realised (he did not say why) that there is less of a trend towards IEDs and suicide bombers than the figures show, and that the Taliban are still conducting formation attacks.

Professor Farrell was asked about Dr Giustozzi’s report, in which he said, at paragraph 4, that there are more military casualties in Afghanistan than Iraq, and more civilian casualties in Iraq, and that indiscriminate terrorist attacks are not practised in Afghanistan. Mr Palmer said the respondent does not go quite as far as that, but asked Professor Farrell if he had any comment. His response was that Dr Giustozzi’s view is not inconsistent with the evidence which he had put forward. He said the direct risk to civilians is low, but the indirect risk is great because of the last thirty years of conflicts. He was also referred to paragraph 5 of Dr Giustozzi’s report, where the latter said:

“Like in Iraq, the risk to civilians is not evenly distributed around the territory of the country. It is highest in the southern provinces and along the highways going from Kabul to the provinces of the south and south-east. Significant levels of risk then exist in the south-east and in the east. The risk is lower in the north, north-east and in the central highlands. In Kabul City the risk is modest: there have been bloody attacks but the casualty rate among the city’s 5 million inhabitants is rather low.”

Asked if he had any comment, Professor Farrell said that he broadly agreed. The view was perfectly reasonable in relation to the direct risk to civilians. It was put to him that the indirect risks are not evenly distributed, and are relatively light in areas where aid can reach. He agreed with that, and said it was more difficult to know how the pattern will develop. He reminded us that the UN Secretary General’s report to the Security Council spoke of the worsening situation throughout Afghanistan. He said that, to some extent, the degree of indirect risk reflects the reach of ISAF. There is a delay before aid gets into areas where ISAF have deployed because, when they first arrive, they come against enemies in the short term. He accepted that increasing troop numbers will help overall, but asked rhetorically how long it will take and whether the United States will remain in Afghanistan.

He was then asked about the last sentence in his report, in which he said the direct risk to civilians of violence from insurgents, terrorists and criminal gangs will remain very high. Professor Farrell said that he would add warlords “into the mix” as well as the Afghan security forces, and that in referring to direct risk he was talking of injury and fatality. It was put to him that criminal gangs have not featured in his report. He said that was really included in his reference to warlords. He said that in Nangarhar the governor is a particularly unsavoury character. Asked what risk that presented, he said that, with such a person as a governor, “What can the population expect of its local government?” Asked whether the criminal gangs caused a direct risk of violence he said the point is that, whilst in parts of Afghanistan there are attempts to restore law and order, in other parts there is no protection at all. As an impression, he said risk from criminal gangs will be high. Asking about indirect risk associated with chronic state failure, Mr Palmer suggested that would need to be looked at by area. Professor Farrell agreed, saying that it is not the problem in Kabul. He was not sure whether it was the case in Jalalabad, although it is generally in Nangarhar, which is described as one of the most challenging provinces, and where state failure is at its highest. Mr Palmer suggested that risks in Nangarhar were lower than in other provinces, and Professor Farrell said that according to the new table, a comparison of IED events by province and regional command, Nangarhar had the second highest number (after Khost). Of the 89 events, 42 were successful (i.e. detonated) and 47 were not. He accepted that the numbers of devices did not put the level of violence into an exceptional category, although he said the general climate is very hostile to basic necessities and to sustainable living. He said that, in Nangarhar, IEDs form a direct risk to people, whereas the indirect risk there was associated with chronic state failure. Asked whether he could point to any evidence to show that even indirect risk was extreme in Jalalabad, he said he could not, but there are problems throughout Afghanistan.

Asked by the Tribunal whether there was a correlation between the number of EDPs outside Afghanistan and the level of violence; and whether one could infer that the numbers were directly as a result of what happens on the ground now, or at the time when they left, Professor Farrell said it was possible it was a reflection of what happened when they left, but he was not an expert on “refugee flows”. He went on to say that, when comparing food security with the number of EDPs, the position did not appear to have improved on the ground.

On the second day of the hearing, the appellant handed in an Updated Version of his Extracts from Country Evidence; a document entitled Afghanistan Index – Tracking Variables of Reconstruction and Security in Post-9/11 Afghanistan (Brookings, July 15, 2009); and a UN Report of the Security Council mission to Afghanistan, 21 to 28 November 2008. The respondent served the UNHCR Eligibility Guidelines for Assessing the International Protection Needs of Asylum-Seekers from Afghanistan. We have taken these materials into account, together with the rest of the evidence. They do not materially differ from the general thrust of the other evidence. Besides the submissions recorded, we observe that paragraphs 36 and 37 of the Security Council’s Report describes a deteriorating humanitarian situation, “resulting from a combination of natural disasters, lack of Government capacity to prepare and respond, and insecurity”, aggravated by “the deficit in food production”. Some 40,000 Afghans die every year “as a result of hunger and poverty, 25 times more than die as a result of the violence”. On page 43 of the UNHCR’s Eligibility Guidelines, it is said that “due to the fluid and volatile nature of the conflict, lack of comprehensive monitoring and reporting from all conflict areas and variations in terms of numbers of civilian casualties reported, UNHCR is not in a position to designate specific conflict areas of Afghanistan in which there is a serious and indiscriminate threat to the life, physical integrity or freedom of Afghans as a result of generalized violence or events seriously disturbing public order”. It is suggested that claims based on a fear of serious and indiscriminate harm arising from the armed conflict should be assessed individually on their merits.

We have found that in the Afghan context, in general, the indirect consequences of the armed conflict described in paragraphs 66 to 70 above do not contribute to the individual risk to people generally, because of the remoteness of the consequences and the requirement for a causal link. For that reason we do not need to deal with much of the background material which has been placed before us, and which goes to such matters as life expectancy, food poverty and other adverse impacts of war on human security short of serious injury and death. Even then, because of the reference in Article 15(c) to indiscriminate violence, we are not concerned with those who are directly targeted in assassination attempts, or those who are accidentally harmed by combatants, but where neither party is acting in disregard of the safety of civilians, for example by firing at insurgents taking shelter amongst civilians.

The respondent’s case is that the civilian casualty figures in Afghanistan fall very far short of demonstrating the high levels of indiscriminate violence necessary to meet the test set out in Elgafaji. Mr Palmer referred to the civilian casualty figures for 2008 which showed that 2,118 deaths had arisen that year, of which 828 had been caused by government forces and 1,160 by AGEs. He argued that the UN Secretary General’s Report of 10 March 2009 and the UNAMA Annual Report 2008 show that the casualties from pro-government forces’ activity arose as a result of (a) a limited number of larger incidents and (b) air strikes. The UN Secretary General’s Report stated that 68% of casualties caused by pro-government forces had been the result of air strikes.

He then argued that those figures related to a period before a number of recent developments were introduced to tackle them. The first of those is a surge of US troops, and an increase in NATO troops. Second, there has been the important announcement by General McChrystle, the United States Commander in Afghanistan, of a new directive which restricts the use of US air strikes in areas where civilians are at risk. The General indicated to the US Senate Armed Forces Committee that the struggle might be at a critical point in gaining the support of the Afghan people. General McChrystle has also issued two directives with measures aimed at reducing the number of civilian casualties and establishing a process to improve the collection of relevant data. They followed a promise by the American Ambassador that coalition forces would change their tactics in order to prevent civilian casualties in the future. In their skeleton, Mr Palmer and Mr Blundell suggested that these developments show that civilian casualties caused by pro-government forces cannot be described as being caused by indiscriminate violence. In the absence of evidence that the reality on the ground has changed, we would not go so far as to say that no civilian casualties caused by pro-government forces can be so described, but it is a step in the right direction and many of the civilian casualties caused by pro-government forces will not have been caused by indiscriminate violence. The respondent cited the UNAMA Report as demonstrating that the new directives from the US military have already yielded results. The report shows that, apart from one incident when an air strike erroneously targeted a wedding celebration, killing 37 civilians, there was a reduction in civilian casualties caused by pro-government forces in October and December 2008. That strike on a wedding party is one reason we consider the submission - that no civilian casualties caused by pro-government forces can be described as having been caused by indiscriminate violence - is an overstatement.

Mr Palmer asserted that there is evidence that the tactics employed by AGEs have changed in a way that will lead to fewer civilian deaths. He said that the Taliban have begun to target military individuals more specifically, and have moved away from face-to-face confrontation. It is said that marks a shift from attacks on civilian targets. It is reported, both in the Newhaven Register for 3 January 2009, and an article “Possible Change in Taliban Tactics” in the Daily Times for 8 June 2009, that their tactic is increasingly to use explosives and bombings against the military, and long range marksmen to fire on US troops. Mr Palmer argued that the recent developments show that the figures for civilian casualties for 2008 should not be viewed as casualties arising from indiscriminate violence, both because of the coalition forces’ attempts to reduce civilian casualties, and the Taliban’s move away from some of its traditional techniques.

In a further analysis of the figures, the respondent’s skeleton argument asserts that only a small subset of civilian casualties constitutes true victims of indiscriminate violence. The UNAMA Report for 2008 suggests that, of the 2,118 civilian casualties for 2008, 1,160 were caused by anti-government elements. Of those, 207 were the victims of assassinations. By definition these are the victims of targeted violence which cannot be said to be indiscriminate, save possibly to the extent that they have been targeted irrespective of their civilian status. The skeleton suggests that, particularly in assessing risk to the appellant, targeted victims should not be counted as victims of indiscriminate violence, because the appellant himself is not at such a risk, as he has no features that would put him in an enhanced risk group. It is also suggested that the 725 victims of suicide and IED attacks must contain a proportion that were “caught in the crossfire”. The skeleton argument accepts that there are examples of such techniques which clearly do amount to indiscriminate violence, such as the suicide bomb attack referred to by UNAMA where a prominent tribal leader was targeted with the result that, not only were 12 of his auxiliary police colleagues killed, but also 67 civilians. Similarly, a bomb was detonated in Khost, in December 2008, as a group of schoolchildren were walking past, killing 5 of them. However, not all the documented attacks belong in that category, says the skeleton argument. Among those that would not are a suicide car bomb attack on a convoy of foreign troops on 31 January 2009 which clearly targeted against military personnel, but in which two civilians were also killed; an attack on a convoy of Afghan police which killed one child, as well as four police officers, in April 2009; and an attack on a security firm convoy on 4 May 2009, which killed two civilians.

Finally, of the 164 victims in 2008 that were killed by “other AGE tactics”, the skeleton asserts that it cannot be assumed that these were victims of indiscriminate violence as not all rocket attacks or ground engagements which affect civilians would amount to indiscriminate violence. In conclusion, the respondent argues that only a minority of the total civilian casualties of 2,118 can properly be said to be relevant victims of indiscriminate violence for the purpose of establishing a level of risk which the appellant may face. The figures do not demonstrate the very high level of risk for an individual without enhanced risk factors which may render him a target for insurgents.

In their skeleton, Counsel for the respondent suggested that the situation in Jalalabad is such that security incidents continue to take place there but the objective evidence does not suggest that the risk of such events taking place is high enough to amount to a general Article 15(c) risk. They argue that there is no satisfactory evidence that Jalalabad is any more dangerous, for an ordinary civilian, than any other part of Afghanistan. Such objective evidence as there is about isolated security incidents does not suggest they amount to an Article 15(c) risk.

That the trend of reducing casualties has continued into 2009 was confirmed by NATO on 22 April 2009. EarthTimes.org reported that Western and Taliban-led forces had killed almost 40% fewer Afghan citizens in the first three months of the year than in the same period in 2008. It was said by Counsel that this reflects, at least in part, increased efforts to reduce the body count. According to NATO, Taliban-led fighters killed four times as many Afghan civilians as did international forces in the first quarter of the year and that proportion has not changed from the previous year. They accepted that a further internet report from Afghanconflictmonitor.org suggested, on 20 July this year, that figures used by the United Nations and the Afghanistan Independent Human Rights Commission may well represent a substantial undercount but argue that this does not mean that the underlying trend they have demonstrated is wrong.

Before moving on to the appellant’s submissions we should comment that one difficulty with the figures that we have been provided with, is that they tend to be Afghanistan-wide. It is apparent from the province by province analysis which appears in the Brookings Institution Afghanistan Index that the incidents are not evenly distributed throughout Afghanistan. Another difficulty with the Brookings figures is that, for example, in the most recent report of IED events, there is no indication as to casualties where the events were said to have been successful. There is also no indication as to the division of casualties as between the armed forces and the civilian population. We should add that a large amount of the statistical information we have seen deals with military and police casualties. In our judgment, that does not help, to any significant extent, to demonstrate the risk to civilians. Whilst some incidents will undoubtedly take place in areas where there are civilians present, others may well occur in isolated or unpopulated areas.

For the appellant, Mr Symes had attached to his skeleton argument extracts from the country evidence that had been produced by the IAS Research Unit. He provided an updated version on the second day of the hearing. The extracts deal with three sections of the bundles of documents. The first with humanitarian issues and health, the second with violence in Jalalabad and Kabul, and the third a selection of other passages. In addition, the appellant relies on material supplied on the second day of the hearing but referred to by Professor Farrell, that is to say the Brookings figures on IEDs, and the 2008 UNAMA Report.

The first extract of the country evidence documents deals with humanitarian issues and health. As that is an area we have concluded does not generally fall for consideration under Article 15(c), and in particular not in relation to Afghanistan, for the reasons we have already given, we have not considered those pages (pages 100 to 113 of Tab E of Volume 1) in detail.

We deal with violence in Kabul and Jalalabad later. With regard to the third category of documents, we observe at page 135 in Volume 2 of the appellant’s bundle, details of a briefing to the UN Security Council on 3 July 2009. There it was suggested that if managed well the situation in Afghanistan, ahead of the defining presidential and provincial elections, could become a turning point in the efforts to end the conflict. The spokesman, however, went on to say that he had to be careful not to present a rosy picture of the situation, as the ongoing conflict seriously undermined prospects of progress. The number of security incidents rose above the 1,000 mark for the first time in May, a 43% increase over the same period in 2008. He said it “had been the most intense fighting season so far experienced”. At page 154, the IAS extracts highlight a report dated 23 June 2005 by the Secretary General on the situation in Afghanistan, and its implications for international peace and security. This report also refers to the increase in incidents, saying that the rise is due to increased fighting in the traditional conflict areas in the South and East, but also that insurgent activities in previously stable areas, particularly in the North, have expanded. On a more positive note, it was reported that as a result of improved coordination between the Afghan National Army, the Afghan National Police Force and the National Directorate of Security there has been a sharp decrease in security incidents in Kabul and the surrounding provinces. The report refers to the free movement of unarmed civil servants being adversely affected, and makes mention of the continuing deliberate targeting of government officials and employees, religious scholars, civilian contractors and the government and the aid community. It said that, compared with the first four months of 2008, the reporting period had been marked by an increase in assassinations, abductions, incidents of intimidation and the direct targeting of aid workers, including the United Nations staff, although the number of aid workers killed had dropped sharply, with only one case being reported during the first four months of 2009.

In assessing the evidence, the one thing which struck us particularly was Professor Farrell’s assertion that the number of civilian fatalities directly caused by both sides to the conflict in Afghanistan (including those assassinated by the Taliban/al Qaeda) was low in comparison with conflicts of a similar size elsewhere. This emerges in particular from the questions put to Professor Farrell regarding page 17 of his report. Whilst it is apparent that any assessment of risk to civilians needs to cover not only those casualties, but also those who are injured as a result of intimidation by insurgents (which the table on that page does not cover and as to which no reliable data was presented), Professor Farrell was nevertheless clear that the current conflict in Afghanistan cannot be said to involve a high level of civilian casualties (albeit that he urged us to take account of what he considered were the conflict’s indirect effects). We conclude that the number of direct victims of indiscriminate violence, arising as a result of the armed conflict, does not demonstrate that the appellant, upon whom the burden of proof lies, has established that there is such a high level of indiscriminate violence that there are substantial grounds to establish that he would, solely by being present in that country, face a real risk which threatens his life or person. We reach this view bearing in mind what the Court of Justice had to say. The appellant has not shown that incidents of indiscriminate violence are happening on so wide a scale, and/or in such a way, as to pose a serious threat of real harm. So far as indirect effects of violence are concerned, we have already explained why we do not consider that these can be said to fall within the scope of Article 15(c).

We acknowledge that the material provided, in particular that referred to by Mr Symes, deals in a qualitative way with various incidents, not least the unfortunate bombing by the United States of a wedding party in November last year. That incident has clearly affected the statistics that would otherwise demonstrate an improvement since the change in tactics announced by General MaChrystle.

We have considered the appellant’s evidence concerning safety in Nangarhar (in which Jalalabad is situated), which is to be found in Section J of the IAS extracts of the evidence. This qualitative evidence deals with a number of incidents. The first was that on 18 May 2009, in eastern Nangarhar, four Afghan National Army soldiers and their companions were arrested for shooting three shopkeepers at a plaza in Jalalabad City. Mr Palmer made the point that, if anything, this showed that there was a degree of protection available from the state.

On 11 May 2009, Voice of America referred to eighteen people, including eleven construction workers, being killed by bomb and suicide attacks the previous Sunday. The source article, found on page 551 of the appellant’s bundle, shows that seven of those killed were in Helmand and three in Zabul, only eight being in Nangarhar. Those eight were killed when their vehicle hit a roadside bomb and of course it is not known whom that was intended to catch.

On 21 March 2009, there was reference to seven civilians and a police officer being killed in eastern Nangarhar. The source document, on page 564 of the appellant’s bundle, says more and illustrates the difficulty of deciding whether incidents should properly be described as involving indiscriminate violence. In that incident, a suicide bomber wanted to attack a convoy of US army vehicles. As it approached the convoy, the car he was driving hit a tractor trailer and exploded.

On 17 January 2009, it was reported that a suicide bomber, targeting foreign troops in eastern Nangarhar, killed one civilian and wounded six people, including three police. Again, it is not clear that that can be properly described as indiscriminate violence.

Finally, the same incident is reported twice. In November 2008, a suicide bomber rammed a vehicle into a US convoy as it drove through a busy market. The reports are not entirely clear on where this occurred, one saying it was in Jalalabad. Eleven died, including one soldier, with possibly 74 others wounded. It seems to us that this incident is properly described as involving indiscriminate violence because the incident occurred in a crowded market.

It is very difficult, from reading a number of qualitative reports concerning various incidents occurring in different parts of a country, to get a reliable feel for what is really going on. Many of the incidents are reported more than once, and the political stance of those reporting the incident is not always clear. Nobody is suggesting that the situation in Afghanistan is anything but a very long way short of ideal but, in order to assess the risk in this case it was necessary to analyse the situation quantitatively. That has been possible because of the detailed statistics about the numbers of attacks, and the level of casualties. It has been possible to get a good indication of the proportion of civilian casualties which can be attributed to indiscriminate violence. When that is done, and to the extent that we have already demonstrated, the numbers of civilians killed by indiscriminate violence turns out to be a great deal less than might otherwise have been expected.

Professor Farrell said in his evidence that one view was that for each death arising from indiscriminate violence, there could be something like an average of eight wounded. Obviously that figure is very variable but, even if it were correct, it is not possible to know the proportion of those casualties which represents the very seriously injured, or how many suffer little more than cuts and bruises. It was an interesting observation, but not one that we found particularly helpful in our analysis. Even if we assume that a proportion of that figure was seriously injured, then the numbers of casualties involved would not elevate the risk sufficiently to bring the general situation within Article 15(c).

What we have just said about the statistics is true for Afghanistan as a whole. The limited amount of evidence available about the incidents on a province by province basis, either in Nangarhar, or more particularly Kabul, does not suggest that the risk is at the high level required to satisfy Article 15(c). It therefore follows that the appellant will not be at real risk of serious harm as defined in Article 15(c) in his home area or in Kabul.

The incidents cited by Mr Symes are further evidence of continuing violence, but do not themselves suggest that the more quantitative assessment carried out and based on the statistics produced by Professor Farrell, is wrong. They illustrate the day to day hazards of living in Afghanistan, but do not elevate the risk to such a level that Article 15(c) is engaged.

Having concluded that the appellant would not be at real risk of serious harm as defined in Article 15(c), were he to return to Jalalabad, the next question is whether he can safely get there. We have evidence about the road between Kabul, to where it is agreed he would be returned, and Nangarhar. This was dealt with by the respondent’s counsel in their skeleton argument. They refer to a report by the International Council on Security and Development (“ICOS”) (formerly the SENDIS Council) that the Jalalabad road is not safe after the Sarobi Junction. In the bundles, there are other reports which describe the highway in more positive terms, saying that it has been improved in terms of resurfacing since the overthrow of the Taliban. The respondent argues that none of the other reports refer to serious security incidents of an extent that would give rise to an Article 15(c) risk. It is of course for the appellant to establish that he would not be able to access Jalalabad safely. For this purpose, the appellant would in our judgment have only to show that he would be at real risk of serious harm from any cause, not limited to that in Article 15(c).

The ICOS Report says little more than was quoted in the skeleton argument. The report has been criticised by a NATO spokesman who described figures contained in it, relating to the Taliban presence in Afghanistan, as “not credible at all”. Observations about the ICOS methodology are to be found in a report from RFE/RL (Radio Free Europe/Radio Liberty) on 8 December 2008. The report deals with the headline findings of ICOS but then reports the NATO spokesman as saying that NATO and the Afghan Government reject the report. The report was said by the Afghan Government to have used questionable methodology, to be conceptually confused and to have misinterpreted the sporadic terrorising, and media orientated activities, of the Taliban.

The other evidence about the Kabul to Jalalabad road is found between pages 347 and 371 of Volume 4 of the evidence. The author of a report dated 30 June 2009 headlines his article “A Journey on the Treacherous Road to Kabul” but his headline is not borne out by the text. The author was startled by an insurgent attack on an American Humvee outside the entrance to the Ministry of Agriculture in downtown Jalalabad, but that was clearly targeted violence not involving the highway. The author then set off towards Kabul and encountered a convoy of vehicles carrying mine-resistant ambush protective vehicles (“MRAPs”) towards one of the US bases. One MRAP had come off its flat bed trailer while rounding a steep mountain curve. He went on to say:

“Afghan drivers are still discovering the dangers of driving too fast on the country’s newly paved roads. The network of roads in this country deteriorated horribly during the past quarter century of conflict. Many Afghan drivers are more accustomed to creeping on bone jarring dirt tracks at speeds of barely 10mph.”

He later came across a traffic jam in the mountains because another MRAP had fallen off a trailer, but said that things should improve shortly because a Chinese company is about to finish “paving” the last stretch of highway through the mountains east of Kabul. Having taken to walking, to avoid the traffic jam, the author then found a gang of dozens of motorcycle taxi drivers offering to ferry travellers back to Kabul. Many were revving motors and doing wheelies. Some yodelled, others performed tricks, either driving the bikes using their feet or lying on their stomachs on the seats of bikes as they raced through the mountains. There is a less subjective report on the highway by Eng Consult Limited which shows a high level of investment in the road. A table, at paragraph 7, shows that the road was being used (in both directions) by an average of 5,582 vehicles per day (during the period 24 June to 30 June 2007) and that, of those vehicles, 53% were cars, jeeps, four wheel drives and taxis. The respondent’s skeleton argument suggests that puts into perspective the limited number of incidents reported on that highway.

Such evidence as we have shows that the road between Kabul and Jalalabad is heavily used and subject to considerable investment. That is not to say there will never be incidents on the road but there is no satisfactory evidence that there is a reasonable likelihood of the appellant being caught up in one. There is serious doubt about the ICOS Report, which in any event is extremely brief and devoid of reasoning, to the effect that the road is not safe beyond a particular point. The other evidence provided by the respondent contradicts that, and in the absence of any better evidence from the appellant, he has failed to make out his case in that regard. We do not find that the appellant would be at real risk of serious harm, however defined, when travelling from Kabul to Jalalabad.

Even if we were wrong about the situation in Jalalabad, or access to it, the appellant still has the option of internal relocation to Kabul, to where he will be returned. In this respect the test is not that required by Article 15(c), or any other part of Article 15, but that set out in Article 8 and considered in Januzi v Secretary of State for the Home Department [2006] UKHL 5 and AH (Sudan) [2007] UKHL 49. For that purpose, to succeed the appellant has to show that it would be unduly harsh to expect him to move to Kabul or, in other words, that it would be unreasonable to expect him to do so having regard to all the circumstances. The appellant is a healthy young man with sufficient resources to have made it to the United Kingdom and to have lived here. In PM and Others (Kabul – Hizb-i-Islami) Afghanistan CG [2007] UKAIT 00089 the Tribunal found, following a country guidance case in which Dr Giustozzi gave oral evidence at some length, that it is not unduly harsh or unreasonable in all the circumstances, to expect such a person to relocate to Kabul and that it is not unreasonable to expect him to do so. That issue was considered again, most recently in RQ (Afghan National Army – Hizb-i-Islami – risk) Afghanistan CG [2008] UKAIT 00013, and there was no material change in the assessment.

We asked Mr Symes whether there was anything, in that section of Dr Giustozzi’s report, dealing with general conditions in Kabul, that had not previously been considered by the Tribunal in those earlier cases, or which cast doubt on the continuing reliability of the Tribunal’s earlier findings. He said there was not. In the absence of any evidence to challenge the findings in those earlier country guidance cases we are bound by them and find that it is not unduly harsh or unreasonable in all the circumstances to expect a young man, about whom nothing else is known, to relocate to Kabul.

There is one final matter which we should deal with, as this is a country guidance case. That is the question of enhanced risk to particular categories of individuals. The European Court made it clear in Elgafaji that where a person comes within a group of people for whom there is an enhanced risk, the degree of indiscriminate violence does not need to be as high as it would otherwise have to be in order to invoke Article 2 or Article 15(c). We have already observed that the ultimate test is that of real risk of serious harm. We have not heard much evidence about enhanced risk categories, and that is not an issue we have had to consider in relation to the appellant. It was accepted by counsel for the respondent that those who could be perceived as collaborators may be considered to be in such a category. That may include teachers, local government officers and government officials. The concept of a group of people at enhanced risk of indiscriminate violence is not an immediately obvious one. The difficulty concerns the use of the word “indiscriminate”, but the answer is partly contained in QD and AH, which considered the “individual risk of indiscriminate violence”. The way in which an enhanced risk might arise for a group can best be demonstrated by example. If, say, the Taliban wanted to make a point about teachers continuing to teach girls, it may resolve to kill a teacher. It would not be any specific teacher but one who came into their sights. A teacher is of course not a combatant and an attempt to kill the first teacher they came across could be argued to demonstrate that teachers were then at enhanced risk of indiscriminate violence. Another possible example could be disabled people. If a bomber, or sniper, were to walk into a crowded marketplace, the public may well flee. A man with only one leg would move considerably more slowly and arguably as a result would be in a higher risk group than the general public. In view of the paucity of evidence, we cannot give a list of risk categories, and certainly cannot say that any particular occupation or status puts a person into such a higher risk category. We merely record that there may be such categories, and that if a person comes within one, the degree of indiscriminate violence required to succeed may be reduced depending upon the particular facts of the case both in terms of the individual concerned, and the part of Afghanistan from which he comes. It should also be borne in mind that such a person may, depending on the facts, be entitled to refugee status rather than relying on the subsidiary protection offered by Articles 2 and 15 of the Qualification Directive. We emphasise that those examples should not be taken to indicate that teachers, or the disabled, are members of enhanced risk groups, without proof to that effect.

For all the reasons which we have given, and the Immigration Judge having made a material error of law in failing to consider the question of Article 15(c), we substitute our own decision:-