Newly released last month from the European Asylum Support Office (EASO) was a judicial analysis of Article 15(c) of Council Directive 2011/95/EU (the recast Qualification Directive).
The full report — Article 15(c) Qualification Directive (2011/95/EU) - A judicial analysis — runs to 124 pages and can be read here. The 23-page main body of the report (Parts I and II) is excerpted and reproduced below for easy reference.
The EASO analysis aims to put at the disposal of courts and tribunals dealing with international protection cases, a helpful tool for the understanding of Article 15(c).
Article 15(c) defines serious harm as "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 EASO analysis states: "This provision, which by its nature can potentially affect the outcome of many cases dealing with international protection, has not proved easy for judges to apply. Studies show that in different Member States there have been divergent interpretations. The commentary is intended to assist the reader towards an understanding of the QD through the case law of the Court of Justice of the European Union (CJEU) as well as that of the European Court of Human Rights (ECtHR) and relevant decisions of the courts and tribunals of the Member States."
Part I of the report analyses the constituent elements of Article 15(c), including what constitutes 'international or internal armed conflict', and Part II examines how the provision is to be applied in practice. A 'decision tree' in Appendix A sets out the questions courts and tribunals need to ask when applying Article 15(c).
Parts I and II follow below:
Article 15(c) Qualification Directive (2011/95/EU)
A judicial analysis
© European Asylum Support Office 2015
Part 1: The elements
1.1. Real risk of serious harm
1.2. Armed conflict
1.2.1. Internal armed conflict
220.127.116.11. Differentiation between defining internal armed conflict and establishing the level of violence
18.104.22.168. Basis of definition
22.214.171.124. Applying the CJEU definition
126.96.36.199. Must be two or more armed groups
1.2.2. International armed conflict
1.3. Indiscriminate violence
1.3.1. CJEU definition of indiscriminate violence
1.3.2. National Case Law
1.3.4. Typical forms of indiscriminate violence in armed conflicts
1.3.5. The role of targeted violence
1.4. By reason of
1.5.1. Personal scope of Article 15(c): confined to civilians
1.5.2. Approach to definition likely to reject IHL definition
1.5.3. Differentiation between military and non-military
1.5.4. Civilians = all non-combatants?
1.5.5. Does the term 'civilian' exclude all members of the armed forces and police?
1.5.6. Is mere membership of an armed group sufficient to exclude status of a civilian?
1.5.7. Indicators of civilian status
1.5.8. Future-oriented assessment
1.5.9. In case of doubt
1.5.10. Former combatants and forced enlistment
1.6. Serious and individual threat
1.6.1. General risk and specific risk
1.6.2. Concept of a 'sliding-scale'
1.7. [Civilian´s] life or person
1.8. Geographical scope: country/area/region
1.8.1. Identification of home area
1.8.2. Home area as area of destination
1.8.3. Protection against serious harm in area of destination
1.8.4. Internal protection
Part 2: Application
2.1. Résumé: Holistic approach
2.2. Assessing the level of violence — a practical approach
2.2.1. Strasbourg case law
2.2.2. National courts and tribunals
2.2.3. UNHCR's position
2.2.4. Conclusions — non-exhaustive list of possible indicators
2.3. Application of the sliding scale assessment
National case law
2.4. Geographical scope: country/area/region
2.5. Internal protection
2.5.1. Article 8 (original and recast QD)
Article 2(f) refers to a 'real risk of suffering serious harm as defined in Article 15'.
Subsidiary protection concerns third country nationals who do not qualify for asylum but for whom substantial grounds have been shown for believing that they would face a 'real risk of suffering serious harm' if returned to their country of origin (see Article 2(f); previously Article 2(e)). As regards the need to show substantial grounds, Member States may consider it is the duty of the applicant to submit as soon as possible all the elements needed to substantiate the application for international protection. On the other hand, it is the duty of the Member State to assess, in cooperation with the applicant, the relevant elements of the application (Article 4(1)). Advocate General Sharpston pointed out in her opinion in Joined Cases A, B and C  that:
'[t]he process of cooperation under Article 4(1) of the Qualification Directive is not a trial. Rather it is an opportunity for the applicant to present his account and his evidence and for the competent authorities to gather information, to see and hear the applicant, to assess his demeanour and to question the plausibility and coherence of that account. The word 'cooperation' implies that both parties work towards a common goal. It is true that that provision allows Member States to require the applicant to submit the elements needed to substantiate his claim. It does not follow, however, that it is consistent with Article 4 of the Qualification Directive to apply any requirement of proof which has the effect of making it virtually impossible or excessively difficult (for example a high standard of proof, such as beyond reasonable doubt, or a criminal or quasi-criminal standard) for an applicant to submit the elements needed to substantiate his request under the Qualification Directive. […] However, when information is presented which gives strong reasons to question the veracity of an asylum seeker's submissions, the individual must provide a satisfactory explanation for the alleged discrepancies.'
The 'real risk' element determines the standard of proof required for eligibility for subsidiary protection . In other words, it denotes the degree of likelihood that the situation of indiscriminate violence will be one that gives rise to serious harm.
To date, the CJEU has not provided a precise interpretation of the notion of 'real risk'. Nonetheless, the Court has confirmed that in relation to Article 15(c), a risk linked merely to the general situation in a country is not, as a rule, sufficient . However, there may be exceptional situations where the degree of indiscriminate violence is of such a high level that an individual would face a real risk solely on account of his presence . In addition, it can be assumed that the 'real risk' standard excludes risks that are at the level of mere possibility or so remote as to be unreal . The degree of risk required under this provision is described in more detail below in Section 1.3 'Indiscriminate violence' and Section 1.6 'Serious and individual threat'.
The 'serious harm' element characterises the nature and intensity of interference with a person's rights; for that interference to be serious it must be of sufficient severity. Article 15 defines three specific types of harm which constitute the qualification for subsidiary protection. Further, subsidiary protection cannot be granted for any kind of harm, discrimination or breach of rights which an individual may suffer, but only for one of those three types of serious harm which meet the criteria of Article 15(a), (b) or (c).
Bearing in mind the purpose of this document, the following text is focused primarily on serious harm as defined in Article 15(c) pursuant to which serious harm consists of 'serious and individual threat to a civilian's life or person by reason of indiscriminate violence in situations of international or internal armed conflict'.
In Elgafaji, the CJEU, whilst not excluding overlap, confirmed that the harm defined in Article 15(c) covers a more general risk of harm than Article 15(a) and (b) . According to this judgment, what is required is a 'threat … to a civilian's life or person' rather than specific acts of violence. Furthermore, if the level of indiscriminate violence is sufficiently high, such a threat can be 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 . The individual elements of this definition are thoroughly elaborated upon in the following parts of this document.
In addition, the types of harm referred to in the categories of Article 15 may to a certain extent overlap from a factual perspective not only with each other but also with acts of persecution as defined by Article 9 . In such a case, it is necessary to bear in mind the priority of granting refugee status provided that the other conditions of Article 2(d) are met. The CJEU has stated that Article 15(b) corresponds in essence to Article 3 ECHR .
The phrase used in Article 15(c) is 'international or internal armed conflict'.
The meaning of this term was clarified by the CJEU in Diakité. At paragraph 35 the Court confirmed that
'[…] on a proper construction of Article 15(c) of Directive 2004/83, […] an internal armed conflict exists, for the purposes of applying that provision, if a State's armed forces confront one or more armed groups or if two or more armed groups confront each other. It is not necessary for that conflict to be categorised as 'armed conflict not of an international character' under international humanitarian law; nor is it necessary to carry out, in addition to an appraisal of the level of violence present in the territory concerned, a separate assessment of the intensity of the armed confrontations, the level of organisation of the armed forces involved or the duration of the conflict.'
This construction achieves two things:
Short definition — it furnishes a short definition of internal armed conflict (as existing where 'a State's armed forces confront one or more armed groups or if two or more armed groups confront each other' .
Rejection of International Humanitarian Law (IHL)-type approaches — it expressly rejects two alternative approaches to definition. The approaches rejected are described as an IHL approach and an approach which considers that an internal armed conflict only exists if the conflict is of certain intensity, features armed forces with a level of organisation or has a certain duration. Since the latter is essentially an IHL approach, it is reasonable to assume the CJEU rejects 'IHL-type' approaches .
Of particular importance to the CJEU in Diakité was that courts and tribunals keep separate:
• assessment of the existence of an armed conflict; and
• assessment of the level of the violence.
The existence of an armed conflict is a necessary but not a sufficient condition for Article 15(c) to be engaged. In relation to general risk to civilians , Article 15(c) will only be engaged if the latter assessment discloses that the armed conflict is characterised by indiscriminate violence at such a high level that civilians as such face a real risk of serious harm. Thus at paragraph 30 of Diakité, the CJEU observed:
'Furthermore, it should be borne in mind that the existence of an internal armed conflict can be a cause for granting subsidiary protection only where confrontations between a State's armed forces and one or more armed groups or between two or more armed groups are exceptionally considered to create a serious and individual threat to the life or person of an applicant for subsidiary protection for the purposes of Article 15(c) of Directive 2004/83 because the degree of indiscriminate violence which characterises those confrontations reaches such a high level that substantial grounds are shown for believing that a civilian, if returned to the relevant country or, as the case may be, to the relevant region, would — solely on account of his presence in the territory of that country or region — face a real risk of being subject to that threat (see, to that effect, Elgafaji, para 43).'
The CJEU describes its definition of armed conflict as being based on 'its usual meaning in everyday language, while also taking into account the context in which it occurs and the purposes of the rules of which it is part' (Diakité, para 27). We have already noted that thereby the Court makes clear that a specific EU approach to interpretation must be adopted in respect of Article 15(c).
Plainly the CJEU wishes to underline that courts and tribunals are not to seek to deny Article 15(c) protection on the basis that the armed confrontations taking place do not meet the threshold required under IHL or any comparable extrinsic body of standards.
In paragraph 17 of Diakité, the CJEU described the first question it had to answer as having two parts: (i) if the assessment of whether an internal armed conflict exists is to be carried out on the basis of the criteria established by IHL; and (ii) 'if not, which criteria should be used in order to assess whether such a conflict exists […]'.
The CJEU gives a clear no to (i), but as to (ii) it does no more than offer its very short everyday language definition. As a consequence, it is left to courts and tribunals to unpack and/or operate this definition in practice. The CJEU definition is clearly broader than the IHL definition and could include, for example, armed confrontations flowing from the drug wars in some Latin American countries . Accordingly, depending on the country situation it may still be necessary for courts and tribunals in certain circumstances to decide whether there is an armed confrontation in the sense described by the Court. For example, riots and insurrections wholly or mainly lacking any use of arms would not appear to qualify. Use of arms alone may not be enough unless there is use of them within or by armed groups. The existence of armed groups alone may not be enough, for example, if such groups do not in practice use arms. There would also need to be evidence of confrontation (i.e. fighting) between them or between an armed group and State forces.
The CJEU's definition would appear to exclude a situation where there was only one armed group confronting the general populace, although Advocate General Mengozzi in his opinion in Diakité (as with the English Court of Appeal in QD (Iraq))  advocated that this too should be covered. However, such a situation may be relatively rare.
In Diakité, the CJEU did not seek in terms to define 'international armed conflict' but, pari passu with its reasoning in respect of 'internal armed conflict', it would seem that this term must also be given its usual meaning in everyday language and so must be one that does not impose an IHL threshold. Nonetheless it is likely (as in IHL) that there may be situations where a country is in a state of an internal and an international armed conflict at the same time.
'Indiscriminate violence' refers to the source of the specific type of serious harm identified in Article 15(c). As this provision aims to offer (subsidiary) protection to those civilians who are suffering from the consequences of an armed conflict, the meaning of 'indiscriminate violence' must be interpreted in a broad way.
The protection needs of a specific civilian population in a country or one of its regions should not be determined by a narrow approach to defining the terms 'indiscriminate' or 'violence', but by a careful and holistic assessment of the facts coupled with a close and exact analysis of the level of violence, as regards the nature of the violence and its extent.
In its judgment in Elgafaji, the CJEU has held that the term 'indiscriminate' implies that the violence 'may extend to people irrespective of their personal circumstances' .
The CJEU has highlighted the 'exceptional situation' needed for Article 15(c) to apply to civilians generally. In Elgafaji at paragraph 37, the Court made clear that, for this to be the case:
'[…] the degree of indiscriminate violence characterising the armed conflict taking place … [must reach] 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 in Article 15(c) of the Directive.'
Since the Elgafaji ruling, national courts and tribunals, rather than trying to define the concept further, have sought to identify indicators of its nature and extent (see Part II Section 2.2. below). The United Kingdom Upper Tribunal (UKUT) has stated that bombings or shootings:
'can properly be regarded as indiscriminate in the sense that, albeit they may have specific or general targets, they inevitably expose the ordinary civilian who happens to be at the scene to what has been described in argument as collateral damage. The means adopted may be bombs, which can affect others besides the target, or shootings, which produce a lesser but nonetheless real risk of collateral damage .'
As to general targets, the UKUT gave the example of the explosions of bombs in crowded places such as markets or where religious processions or gatherings are taking place . The German Federal Administrative Court (FAC) interpreting the Elgafaji judgment came to the conclusion that it is not necessary to determine whether the acts of violence constitute a breach of international humanitarian law, because the notion of violence used in the QD was a broad one . There has been considerable discussion in the national case law as to what extent indirect effects of indiscriminate violence should be taken into consideration.
The French Council of State has referred to attacks and abuses against the civil population and to forced displacements as possible characteristics of indiscriminate violence . Such characteristics were satisfied where an applicant had to travel through regions of Afghanistan affected by such violence ; the assessment did not require analysing the nationwide general situation but the regions concerned .
In two judgments, the Administrative Court of the Republic of Slovenia put forward the following factors that should be taken into account in assessing the level of violence: battle deaths and injuries among the civilian population including possible temporal dynamics of numbers of deaths and injuries, number of internally displaced persons, basic humanitarian conditions in centres for displaced persons, including food supply, hygiene and safety and the degree of 'State failure' to guarantee basic material infrastructure, order, healthcare, food supply, drinking water. The Administrative Court pointed out that the protected value in relation to Article 15(c) is not mere 'survival' of asylum seekers, but also a prohibition against inhuman treatment . The Slovenian Supreme Court decided that these factors are 'legally relevant' .
To similar effect, UNHCR understands the term 'indiscriminate' to encompass 'acts of violence not targeted at a specific object or individual, as well as acts of violence which are targeted at a specific object or individual but the effects of which may harm others' .
The nature of the violence can be one major factor in determining whether the violence appears to be indiscriminate. Examples of such acts of indiscriminate violence might include: massive targeted bombings, aerial bombardments, guerrilla attacks, collateral damage in direct or random attacks in city districts, siege, scorched earth, snipers, death squads, attacks in public places, lootings, use of improvised explosive devices etc.
The more the assessment of the nature of violence indicates that the person concerned has been or would be a victim of a targeted assault, the more alert courts and tribunals should be to whether such a person is in fact eligible for refugee protection, not subsidiary protection. But in any event there is no reason to leave targeted violence out of the equation when analysing the level of indiscriminate violence in the relevant area or region of the country. Targeted violence encompasses both specific and general targeting: some violence, albeit targeted, can harm civilians in significant numbers .
Further analysis of how to go about assessing the level of indiscriminate violence is given in Part II at Sections 2.2 and 2.3.
Subsidiary protection under Article 15(c) is granted to any person in respect of whom substantial grounds have been shown for believing that the person concerned, if returned to his or her country of origin, would face a real risk of a serious and individual threat to his or her life or person by reason of indiscriminate violence. A crucial element in considering causation will be the level of such violence . Given the broad definition of indiscriminate violence, the requirement of a causal nexus should not be applied in narrow fashion. The effects of indiscriminate violence can be indirect as well as direct. Indirect effects of the acts of violence such as a complete breakdown of law and order arising out of the conflicts should be considered to a certain extent as well.
Should criminal acts that are the result of a breakdown of law and order and other indirect effects of indiscriminate violence be regarded as constituting indiscriminate violence within the meaning of Article 15(c)?
In 2008, the German FAC decided that criminal violence, which is not committed by one of the parties to the conflict, should be taken into account only when assessing the nature of the serious and individual threat to life or person . According to the FAC, 'the general threats to life that are purely a consequence of an armed conflict — for example, through a resulting deterioration in supply conditions — cannot be included in the assessment of the density of danger'  and therefore do not constitute a threat within the meaning of Article 15(c). The UKUT acknowledged in 2010 that general criminality which causes harm of the necessary degree of seriousness could be a consequence of an armed conflict where normal law and order provisions are disrupted. A serious breakdown of law and order permitting anarchy and criminality occasioning the serious harm referred to in Article 15(c) can lead to indiscriminate violence in effect even if not necessarily in aim . There must be a sufficient causal nexus between the violence and the conflict, but indiscriminate violence affecting civilians does not necessarily need to be directly caused by the combatants participating in the conflict . Likewise the French Council of State  as well as the Dutch Council of State  have held that indirect effects of armed conflicts should be considered.
Similarly, UNHCR emphasises in this respect that a breakdown of law and order as a consequence of indiscriminate violence or the armed conflict needs to be taken into account. In particular, the source from which the indiscriminate violence emanates is immaterial .
It is not yet foreseeable whether the new and broad approach to the notion of armed conflict taken by the CJEU in Diakité will also lead to wider acceptance that indirect effects of indiscriminate violence can constitute indiscriminate violence within the meaning of Article 15(c).
In logic, being a civilian is a necessary pre-requisite to being able to benefit from protection under Article 15(c) . If an applicant is not a civilian and so falls outside Article 15(c), it will be necessary to check whether refugee eligibility or protection under Article 15(a) and (b) was considered or should be considered, unless the applicant comes within the scope of the exclusion clauses (Articles 12 and 17). Articles 2 and 3 of the ECHR (which are not subject to exclusion clauses) may also be relevant.
Given the wide-ranging nature of the reasons advanced by the CJEU in Diakité for rejecting recourse to IHL criteria to help define armed conflict, it must be assumed that it would not accept an IHL definition of civilian.  Instead, the Court would strive to give the term its usual meaning in everyday language whilst taking into account the context in which it occurs and the purposes of the rules of which it is a part (Diakité, paragraph 27). The fact that even within IHL there is no unanimity as to the definition of this term  might be said to add to the unsuitability of an IHL-based definition.
Dictionary definitions, because they vary widely, offer little help and in any event do not assist with meaning which is in conformity with the objects and purposes of the QD. A simple everyday meaning might be that civilians are those who are not combatants or persons who do not fight; but this is so short as to add nothing of any substance.
From the fact that the CJEU in Diakité clearly contemplates that an armed conflict could arise even without State involvement or the State being a party ('or in which two or more armed groups confront each other'), it can be understood that the term is primarily used to differentiate non-military from military personnel. Military personnel may include both members of a State's armed forces or police and members of rebel or insurgent groups (sometimes called 'irregular fighters').
If recourse was done to the meaning of the term 'civilian' in international human rights law (IHRL)  (which increasingly recognises the complementarity of IHRL and IHL), the term may need to be accorded the same meaning it is given in common Article 3 to the four 1949 Geneva Conventions: 'persons taking no active part in hostilities, including members of armed forces who have laid down their arms or are otherwise hors de combat […]'. The latter part of this statement suggests that no longer taking part in hostilities is not sufficient; a person must take steps to actively disengage .
There are a number of national decisions reflective of this approach. In ZQ (serving soldier)  the United Kingdom Asylum and Immigration Tribunal (UKAIT) pointed out that in IHL the fact that a soldier is off-duty or off-sick does not necessarily result in civilian status. The Tribunal cited the Appeals Chamber of the International Criminal Tribunal for the former Yugoslavia (ICTY) which observed in Prosecutor v Blaskic  at paragraph 114 that: 'the specific situation of the victim at the time the crimes [war crimes or crimes against humanity] are committed may not be determinative of his civilian or non-civilian status. If he is indeed a member of an armed organisation, the fact that he is not armed, or in combat at the time of the commission of crimes, does not accord him civilian status.' In HM and Others the UKUT concluded that the definition of civilian should not include '[a]nyone who involves himself in armed conflict', which includes members of armed forces or police . The International Committee of the Red Cross (ICRC) interprets civilians in non-international armed conflicts as 'all persons who are not members of State armed forces or organised armed groups of a party to the conflict'.
Bearing in mind that the CJEU considers the meaning of key terms requires accounting for the context in which they occur and the purposes of the rules of which they are part (Diakité, para 27), it may be that the term 'civilian' is susceptible of a broader meaning, so as to denote all those who are non-combatants or not fighters or all those who are hors de combat. Thus, for example, in contrast to the apparent position in IHL, a member of the armed forces or police service who would only face a real risk of suffering serious harm whilst off duty in his home region or area might arguably qualify. By reference to the reasoning in Diakité, it may be thought that the Court considered the term was to be given a factual definition rather than be seen to denote a preconceived legal status .
From the CJEU's reasoning in B and D , it would not be correct to simply try and deduce a person's non-civilian status from his membership of an armed group. In B and D, which concerned the application of the exclusion clauses from refugee status in the QD, the Court refused to make automatic assimilations based on either UN Security Council Resolutions or EU instruments adopted within the Common Foreign and Security Policy. At paragraph 89 of B and D, the CJEU said there was no direct relationship between the definition of terrorist acts in this material and the QD 'in terms of the aims pursued'. Therefore, 'it [was] not justifiable for a competent authority, when considering whether to exclude a person from refugee status […] to base its decision solely on that person's membership of an organisation which is on a list adopted outside the framework setup by [the] Directive'. Inclusion in a list or in a stated definition could not substitute for an individual assessment of the specific facts. Neither could 'participation in the activities of a terrorist group […] come necessarily and automatically within the ground for exclusion laid down in [the] Directive'.
Assuming there is no automatic adoption of a definition from IHL or any other extrinsic body of law and that instead in similar fashion as was done in B and D, the CJEU requires a 'full investigation into all the circumstances of each individual case', the following indicators (not necessarily consonant with one another) may be of some assistance:
• A civilian is a person who is not party to the conflict and merely seeks to get on with life notwithstanding the situation of conflict.
• Not being armed may be insufficient to render a person a civilian, who is required to be neutral in the conflict as well.
• Persons who willingly participate in armed groups are unlikely to be considered civilians.
• The definition of a civilian would seem to be intended to exclude participants in a war and therefore covers persons who do not or would not participate actively in hostilities.
• The role of an individual in the organisation should be explored. The question of whether an individual acted (or would act) under duress should be taken into consideration. On the other hand, it should also be taken into consideration that, for example, seemingly civilian political representation in a rebel insurrection could be responsible for decisions resulting in killings.
• Individuals working for military institutions, including military hospitals, may have difficulty in being considered civilians, even if they are obliged to follow military rules of command.
• An individual having a civilian task in the army, such as a doctor, may be considered as a civilian, unless the post carries a military rank.
• Not having a military rank may make it easier for a person to argue de facto civilian status.
• Article 43 on Armed forces of the Additional Protocol to the Geneva Conventions of 12 August 1949 relating to the Protection of Victims of International Armed Conflicts (Protocol 1), 8 June 1977 excludes from the definition of armed forces 'medical personnel and chaplains covered by Article 33 of the Third Convention'. A non-combatant army doctor in a military hospital may be considered to be performing an essentially humanitarian, rather than military, duty promoting the right to life as protected under the Charter and the ECHR .
• Visual perception is one of the criteria for recognising civilians and differentiating them from combatants. For the determination of the status, it is necessary to examine only the person's assignment as a non-civilian and the question whether the individual could be identified as a non-civilian upon return.
It should be borne in mind that, in assessing all international protection claims, courts and tribunals are primarily concerned with hypothetical risk on return, i.e. what an applicant's situation will be if returned to his or her country of origin. Questions about whether someone was previously a civilian or combatant/fighter will not necessarily establish whether they will be (or be perceived to be) a civilian or combatant/fighter on return.
If a fact-specific approach is taken to whether a person is a civilian (i.e. would be a civilian on return) then importance should be placed on the principle that, to quote from Article 50 of Additional Protocol I, which is headed 'Definition of civilian and civilian population' at subparagraph 1: 'in case of doubt whether a person is a civilian, that person shall be considered to be a civilian'.
The Belgian Council of Alien Law Litigation  stated that in respect of an applicant who had cooperated with the asylum authorities in trying to establish a claim, the benefit of any doubt should be given in favour of considering such person a civilian.
As regards former combatants (including child soldiers), it should be considered that the purpose of the QD was not to introduce additional exclusion clauses, but to identify persons in need of protection. Consideration of an exclusion clause should normally only be given at a later stage. The French National Asylum Court noted in an Afghan case that a former soldier, who left the Afghan army, can be considered as a civilian .
UNHCR has recommended the following approach:
'In this connection, the term 'civilian' in Article 15(c) should not serve to exclude former combatants who can demonstrate that they have renounced military activities. The fact that an individual was a combatant in the past does not necessarily exclude him or her from international protection if he or she has genuinely and permanently renounced military activities. The criteria for determining whether a person satisfies this test have been defined by the UNHCR Executive Committee .'
This underlines that a former combatant, especially if previously affiliated to the state's armed forces, might still be considered a combatant on return.
The UK Home Office stated in its Asylum Process Guidance on Humanitarian Protection of 15 May 2013 that only genuine non-combatants, i.e. those who are not party to the conflict, qualify for protection under 15(c): 'This could include former combatants who have genuinely and permanently renounced armed activity.'
In general terms, an applicant who has been forcibly enlisted  as a soldier/fighter does not thereby lose civilian status but, as with child soldiers, it would seem that in order to decide the issue, the approach to be adopted should be a fact-sensitive one, similar to that taken by the CJEU in B and D: see above 1.5.6.
Article 15(c) requires an applicant to show a real risk of suffering a serious threat of harm and not necessarily of suffering specific acts of violence. The threat is understood as being inherent in a general situation of conflict and that is, in essence, why this provision covers a more general risk of harm than either Article 15(a) or (b): see Elgafaji, paragraphs 32-34. At paragraph 45 the CJEU stated:
'On those grounds the Court (Grand Chamber) hereby rules: Article 15(c) of Council Directive 2004/83/EC …, in conjunction with Article 2(e) thereof, must be interpreted as meaning that:
'— the existence of a serious and individual threat to the life or person of an applicant for subsidiary protection is not subject to the condition that that applicant adduce evidence that he is specifically targeted by reason of factors particular to his personal circumstances;
'— the existence of such a threat can exceptionally be considered to be established where the degree of indiscriminate violence characterising the armed conflict taking place — assessed by the competent national authorities before which an application for subsidiary protection is made, or by the courts of a Member State to which a decision refusing such an application is referred — 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 that threat.'
From the CJEU's analysis in Elgafaji, it is clear that the existence of a serious and individual threat to the life or person of an applicant is not subject to the condition that an applicant adduces evidence of specific targeting by reason of factors particular to his personal circumstances. An applicant can be considered to be at general risk of such a threat if, exceptionally, the level of indiscriminate violence characterising the armed conflict taking place reaches such a high level that substantial grounds are shown for believing that a civilian would face a real risk of being subject to that threat solely on account of his presence in the relevant area or region. Put another way, 'individualisation' necessary to show that the threat is 'individual' may be achieved either by reasons of 'specific risk' factors to do with a person's particular characteristics or circumstances or by the 'general risk' factors arising out of an exceptional situation of a very high level of violence.
Under Article 15(c), whether a person shows either a general risk or a specific risk should not be seen as a dichotomy. Rather the CJEU articulated what is known as 'the sliding-scale' concept, i.e. that:
'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 by him to be eligible for subsidiary protection' (Elgafaji, paragraph 39; Diakité, paragraph 31). The opposite also applies: exceptionally, the level of violence could reach such a high intensity that a civilian would, solely on account of his or her presence on the territory of the affected country or region, face a real risk of being subject to serious harm (paragraph 43). The Court found that this interpretation did not contradict [then] recital 26 of the Directive; as the wording of the latter allows for the possibility of such an exceptional situation. '
By means of the sliding scale concept the CJEU succeeds in balancing individual threat and indiscriminate violence and making clear how the provision is to be operated in a case-specific way.
It can be seen that the CJEU's notion of 'general risk' is similar to the recognition in case law of the ECtHR relating to Article 3 ECHR, of the possibility that an individual can be said to be at real risk of serious harm merely by virtue of being present in a situation characterised by exceptionally high levels of violence. In NA v UK  at paragraphs 115-116 the ECtHR stated:
'115. From the foregoing survey of its case-law, it follows that the Court has never excluded the possibility that a general situation of violence in a country of destination will be of a sufficient level of intensity as to entail that any removal to it would necessarily breach Article 3 of the Convention. Nevertheless, the Court would adopt such an approach only in the most extreme cases of general violence, where there was a real risk of ill-treatment simply by virtue of an individual being exposed to such violence on return.
'116. Exceptionally, however, in cases where an applicant alleges that he or she is a member of a group systematically exposed to a practice of ill-treatment, the Court has considered that the protection of Article 3 of the Convention enters into play when the applicant establishes that there are serious reasons to believe in the existence of the practice in question and his or her membership of the group concerned (see Saadi v. Italy, cited above, § 132). In those circumstances, the Court will not then insist that the applicant show the existence of further special distinguishing features if to do so would render illusory the protection offered by Article 3. This will be determined in light of the applicant's account and the information on the situation in the country of destination in respect of the group in question (see Salah Sheekh, cited above, paragraph 148).'
In Sufi and Elmi v. UK, the ECtHR further clarified that application of this approach would also involve (what we have called) a sliding scale criterion. The ECtHR confirmed, first, that if a risk contrary to Article 3 is established 'the applicant's removal would necessarily breach this article regardless of whether the risk emanates from general situation of violence, a personal characteristic of the applicant, or a combination of the two' (paragraph 218).
One commentator has noted:
'In essence, the Elgafaji 'sliding scale' test does not seem to take much distance from this recent ECtHR jurisprudence, at least on the point of individualisation. Concerning cases of extremely generalised and indiscriminate violence, the test is phrased in similar terms. The CJEU has also made it clear that this situation would be 'exceptional'. Where violence is of lower intensity, both courts require a certain degree of individualisation .'
If there is a 'sliding scale' under Article 3 ECHR, then there must also be one under Article 15(b) . The challenge is how to approach such individualisation in the context of Article 15(c): 'The second challenge stems from the sliding scale test when it comes to the identification of factors particular to the applicant's personal circumstances in cases the violence is of a lower intensity'.  Advocate General Maduro observed that 'when explaining the relevant factors for assessing if a person is individually affected brought as an example their membership of a particular social group [MPSG]' . The MPSG mirrors the 1951 Refugee Convention.
However, if 'personal circumstances' are MPSG or any of the other four grounds of the 1951 Refugee Convention, then the appropriate framework to examine the claim may well be that of the refugee definition .
In any event, the personal circumstances that need to be demonstrated here cannot be limited to the Refugee Convention reasons provided for in the refugee definition; they would in principle appear to encompass factors that would put the person concerned at enhanced risk compared to the rest of the population. It must be recalled that Article 4(3)(c) requires that assessment of an application for international protection must take into account 'the individual position and personal circumstances of the applicant, including factors such as background, gender and age, so as to assess whether, on the basis of the applicant's personal circumstances, the acts to which the applicant has been or could be exposed would amount to persecution or serious harm'.
Whilst therefore the Article 15(c) inquiry is one into specific as well as general risk, the difficulties national courts and tribunals have found in applying the 'sliding scale' suggest that its main utility will be in dealing with claims based on general risk. Claims based on specific risk may very often stand to be resolved under the refugee definition or (if there is no Refugee Convention ground) under Article 15(b) or Article 15(a). It bears repeating that when deciding international protection cases, courts and tribunals must examine first whether a person is eligible for refugee protection and hence application of the 'sliding-scale' under Article 15(c) will only arise if it has been decided an applicant has not established a well-founded fear of being persecuted.
Article 15(c), as stated in the case of Elgafaji , has an additional scope to Article 3 ECHR and, therefore, has to be interpreted independently, but with due regard to fundamental rights as they are guaranteed under the ECHR.
Neither the QD, nor the CJEU in its decisions, has defined the terms 'life or person': two major values of a civilian that are affected by indiscriminate violence in situations of international or internal armed conflict.
By comparing the provisions of Article 15(a) and (b), which indicate a particular type of harm with the provision of Article 15(c), it is obvious that harm thereby defined covers a more general risk of harm .
The harm that could affect an applicant is not restricted to the physical but can also be psychological or mental . The harm could also derive from 'indirect forms of violence, such as intimidation, blackmail, seizure of property, raids on homes and businesses, checkpoints and kidnapping'  which affect a civilian's 'person'. That is the reason why, when examining the risk in case of return, courts and tribunals need to thoroughly examine a wide range of elements to appraise the local situation and conditions.
The question remains open whether the risk to 'life or person' is confined to a real risk of suffering harm that violates non-derogable rights or whether it extends to cover important breaches of qualified rights of an applicant. It was noted in KH (Iraq) at paragraph 101 that:
'[t]his provision, which concerns the focus of the threat, went through five drafting amendments. Dr McAdam (supra at p.75) notes that the original phrase 'life, safety or freedom' was, along with subsequent formulations, based around the concept of freedom ['life or physical integrity or freedom from arbitrary detention'], eventually deleted due to concern by some Member States that it would unduly widen the scope of the Directive .'
Common Article 3 to the 1949 Geneva Conventions uses the phrase 'life and person' (not 'life or person') and in KH (Iraq) it was noted that this phrase is clearly not apt to cover anything to do with civilian objects. The latter is defined in IHL as including the following: 'dwellings, shops, schools and other places of non-military business, places of recreation and worship, means of transportation, cultural property, hospitals and medical establishments and units'. Whilst it is clear from Diakité that key terms of Article 15(c) should not be given an IHL reading, this differentiation would appear to be necessary on any definition.
In KH at paragraph 107 the UKAIT observed a differentiation within Article 3(1) between (a) violence to 'life and person' on the one hand and (c) 'outrages upon personal dignity, in particular humiliating and degrading treatment' on the other. This led the Tribunal to doubt that the material scope of the phrase 'life and person' could extend to threats which amount to inhuman and degrading treatment. The inherent limitation of the concept of 'life or person' within IHL is further indicated by the fact that in Additional Protocol II (by which time it was felt that the protection of civilians should be given a wider material scope) supplementary wording was used to make that protection more expansive. Article 4(2)(a) of the same convention proscribes: 'violence to the life, health and physical or mental well-being of persons, in particular murder as well as cruel treatment, such as torture, mutilation or any form of corporal punishment'. The Tribunal concluded that '[m]indful, however, that 'life or person' is to be given a broad meaning, we would accept that the phrase must encompass the means for a person's survival'. The Administrative Court of Slovenia has held that the protected value in relation to Article 15(c) is not mere 'survival' of asylum seekers, but also a prohibition against inhuman treatment .
It is fundamental to the consideration of Article 15(c) protection that the situation prevailing in the country of return be assessed . However, it is not necessary to decide whether the armed conflict is nationwide; rather the focus must be on the region where an applicant lives (or area of destination) and on determining whether such a person is at risk in that area or the necessary route to it. Article 8 recognises, further, that even if an applicant can establish a real risk of Article 15(c) serious harm in the home area, eligibility for subsidiary protection can only be established if such an applicant is unable to achieve internal protection in another part of the country. The first question, therefore, is whether an applicant is at real risk of serious harm in the home area (or on the way to that home area). If the answer is yes, then the second question is whether serious harm can be avoided by achieving internal protection in another part of the country.
In deciding the location of an applicant's home area as a destination of return, a factual approach is required having regard to matters such as area of last place of residence and area of habitual residence .
When considering risk in an applicant's home area, regard must therefore also be had to whether or not there is an ability to travel to that destination. If not — because of armed conflict affecting the routes that could reasonably be expected to be taken — then such an applicant is to be considered as having demonstrated Article 15(c) risk in his or her area of destination.
The ECtHR had regard to the geographical nature of the conflict in the context of generalised violence in Sufi and Elmi . In national case law on Article 15(c), the German FAC and the French National Asylum Court have found that the assessment does not require analysing the nationwide general situation but the region concerned  including the route to be taken from the point of return to the home area . That is also the consistent position taken by UK courts and tribunals .
It should be noted that when considering whether there exists an Article 15(c) risk in a person's home area, such a risk will only be established if there is no effective protection against it. Article 7  specifies that protection against serious harm must be effective and of a non-temporary nature. Such protection is generally provided when the actors mentioned in Article 7(1)(a) and (b) take reasonable steps to prevent serious harm, inter alia, by operating an effective legal system for the prevention, detection, prosecution and punishment of acts constituting persecution or serious harm, and when an applicant has access to such protection.
If there is an Article 15(c) risk in the applicant's home area (as above), the question will be whether there is a part of the country that is not affected by the conflict to where the individual can reasonably be expected to relocate. This is known as the internal protection (or internal flight, internal relocation) alternative.
Article 8 states:
1. As part of the assessment of the application for international protection, Member States may determine that an applicant is not in need of international protection if in a part of the country of origin, he or she:
(a) has no well-founded fear of being persecuted or is not at real risk of suffering serious harm; or
(b) has access to protection against persecution or serious harm as defined in Article 7; and he or she can safely and legally travel to and gain admittance to that part of the country and can reasonably be expected to settle there.
2. In examining whether an applicant has a well-founded fear of being persecuted or is at real risk of suffering serious harm, or has access to protection against persecution or serious harm in a part of the country of origin 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 in accordance with Article 4. To that end, Member States shall ensure that precise and up-to-date information is obtained from relevant sources, such as the United Nations High Commissioner for Refugees and the European Asylum Support Office.
Recital (27) provides:
Internal protection against persecution or serious harm should be effectively available to the applicant in a part of the country of origin where he or she can safely and legally travel to, gain admittance to and can reasonably be expected to settle. Where the State or agents of the State are the actors of persecution or serious harm, there should be a presumption that effective protection is not available to the applicant. When the applicant is an unaccompanied minor, the availability of appropriate care and custodial arrangements, which are in the best interest of the unaccompanied minor, should form part of the assessment as to whether that protection is effectively available.'
The relevance of internal protection was endorsed by the CJEU in Elgafaji when it stated that 'in the individual assessment of an application for subsidiary protection […] the following may be taken into account […] the geographical scope of the situation of indiscriminate violence and the actual destination of the applicant in the event that he is returned' .
Geographical scope and internal protection are connected principles in that it may be considered implicit in its widest definition that internal protection not only includes protection afforded by third parties  but also self-protection by relocation to a part of the country where the conflict does not exist or where the threat of indiscriminate violence caused by the conflict is less.
Article 8(2) in the recast QD (but not the original — see further below) makes specific reference to access to protection. Article 7 defines actors of protection to include not only State actors but also non-State actors controlling the State or a substantial part of it. The internal protection principle is referable to Article 15 as a whole and it may be considered to have greater application to 15(a) and (b) where individual targeting is the issue rather than 15(c). This is because once a finding of a threat of indiscriminate violence as a result of armed conflict in the home area has been made, the possibility of internal protection within that area being available may not be sustainable because in many situations of armed conflict, there may be little doubt that effective protection is unavailable. 'The capacity of actors of protection to provide protection and indicators relating to State failure' are amongst the indicators for the assessment of the level of violence and serious threat identified by the UNHCR .
The assessment of the situation not just in the applicant's home area but in others parts of the country where it may be possible to find internal protection is therefore essential to the proper consideration of Article 15(c). This assessment with regard to the general circumstances prevailing and the personal circumstances of the applicant requires a thorough appraisal. The QD requires this assessment to take place in accordance with Article 4 (Assessment of elements) and for 'precise and up to date information' to be obtained.
Further analysis of geographical scope and internal protection is given in Part II, Sections 2.4 and 2.5.
In Part I the constituent elements of Article 15(c) were analysed. In this Part, the focus is on how this provision is to be applied in practice.
As noted earlier, the assessment of Article 15(c) requires a holistic approach. Courts and tribunals must take into account a number of elements: armed conflict, civilian's life or person, serious and individual threat, indiscriminate violence, threshold of the violence, geographical scope and the internal protection alternative. There is an interaction between these various elements.
At Appendix A there is a decision tree which aims to help identify the logical order of questions courts and tribunals need to ask when assessing eligibility for subsidiary protection under Article 15(c). In this section, the focus is on the main matters of application that call for further clarification.
The guidance given by the CJEU in Elgafaji  and Diakité  is limited in scope and clearly leaves the matter of how Article 15(c) is to be applied in practice very much to national courts and tribunals. In particular it does not assist the national courts or tribunals in answering the question as to how they should go about evaluating (i) the situation in the relevant area or region of the country so as to assess what is the level of violence and (ii) whether such violence has the effect of creating a real risk of suffering serious harm either to civilians generally or to individuals based on their personal circumstances, or on a combination of the two.
There is no guidance as yet from the CJEU on the criteria for assessing the level of violence in an armed conflict. Courts and tribunals will need to adopt a practical approach to the evaluation of evidence produced in support of the application. Any criteria applied by national courts and tribunals will require a test of practical possibility so as to give effet utile to Article 15(c). At the Member State level, Article 15(c) cases are special because the subject matter is a country where at least parts are in a situation of violence and conflict. As explained in Part 1, courts and tribunals are required to have regard to a number of factors or indicators; in that regard it is important to build on the learning to be found in the case law of the ECtHR and national courts and tribunals.
The approach of the ECtHR to the assessment of the level of violence for the purposes of Article 3 ECHR — so as to decide whether all or most civilians are at real risk of suffering ill-treatment — is set out in Sufi and Elmi at paragraph 241 as follows:
'In the present case the applicants submitted that the indiscriminate violence in Mogadishu was of a sufficient level of intensity to pose a real risk to the life or person of any civilian in the capital. Although the Court has previously indicated that it would only be 'in the most extreme cases' that a situation of general violence would be of sufficient intensity to pose such a risk, it has not provided any further guidance on how the intensity of a conflict is to be assessed. However, the Court recalls that the Asylum and Immigration Tribunal had to conduct a similar assessment in AM and AM (Somalia)  (cited above), and in doing so it identified the following criteria: first, whether the parties to the conflict were either employing methods and tactics of warfare which increased the risk of civilian casualties or directly targeting civilians; secondly, whether the use of such methods and/or tactics was widespread among the parties to the conflict; thirdly, whether the fighting was localised or widespread; and finally, the number of civilians killed, injured and displaced as a result of the fighting. While these criteria are not to be seen as an exhaustive list to be applied in all future cases, in the context of the present case the Court considers that they form an appropriate yardstick by which to assess the level of violence in Mogadishu.
A number of Member State courts and tribunals have adopted a similar approach when assessing the level of violence of armed conflicts for the purposes of Article 15(c). However, there are slight differences in the methods applied as well as in the emphases placed on different indicators.
The UKUT stated that the nexus between the generalised armed conflict and the indiscriminate violence posing a real risk to life or person is met when the intensity of the conflict involves means of combat (whether permissible under the laws of war or not) that seriously endanger non-combatants in a direct or indirect manner . To the Tribunal this meant that focus on the evidence about the numbers of civilians killed or wounded is of prime importance when assessing the level of violence with regard to Article 15(c) . Nonetheless the Tribunal stressed the need for an inclusive approach to assessment of the level of indiscriminate violence. This approach demands an analysis of the level of violence both in quantitative and qualitative terms. A quantitative analysis looks at the numbers of civilians killed or wounded, the number of security incidents etc. A qualitative analysis of the on-going violence has to take account of the impact of threats of violence as well as the physical violence itself, the conduct of the parties to the armed conflict, and long term cumulative effects whenever the conflict is already lasting for quite a while. An inclusive approach that is both quantitative and qualitative should go beyond ascertainment of the figures of civilian casualties — injuries or deaths — and has to bear in mind that population displacement and the degree of State failure are also relevant criteria when assessing the risk of becoming a victim of indiscriminate violence . The UK Tribunal held that even carefully targeted killings that harm no civilians but only combatants, contribute to a climate of fear and insecurity which in an indirect way adds to the intensity of the violence . This is why in the Tribunal's point of view 'it can never be right to attempt some simple subtraction of targeted violence from the overall sum of indiscriminate violence' .
The German FAC stated that an approximate quantitative determination of the total number of civilians living in the area concerned, on the one hand, and on the other hand, the number of acts of indiscriminate violence committed by the parties to the conflict against the life or person of civilians in this region is necessary when assessing the level of violence. Additionally, a general assessment of the number of victims and the severity of the casualties (deaths and injuries) among the civilian population is necessary. To that extent, the criteria for a finding of a group persecution that have been developed under refugee law by the FAC may also be applied accordingly . In addition to the quantitative determination of the level of violence, the approach of the FAC demands a general appraisal of the statistical material with an eye to the number of victims and the severity of the harm (fatalities and injuries) among the civilian population. This general appraisal would in any event also include an assessment of the status of medical care delivery in the territory concerned, on the quality and accessibility of which the severity of incurred bodily injuries may depend, with an eye to the permanent consequences that injuries may have for the victims .
In a case concerning safety in Mogadishu, the Dutch Council of State decided in 2010 that the identification of an exceptional situation in which Article 15(c) shall apply to any individual requires looking beyond the number of deaths and injuries in the area in question to other relevant factors such as internal displacement, refugees fleeing the country and the randomness of the violence .
According to the French National Asylum Court and the French Council of State, the intensity of an armed conflict reaches the threshold of Elgafaji in situations of generalised violence. Forced displacements, violations of international humanitarian law and occupation of territory are also elements to measure the intensity of generalised violence .
UNHCR has likewise urged that courts and tribunals take into account both quantitative and qualitative elements as part of a 'pragmatic, holistic and forward-looking assessment' which 'cannot be reduced to a mathematical calculation of probability' . The organisation draws attention to the caution required when dealing with statistics given the variation in methodology and criteria in the collection of data, the under-reporting of violence, and the relevance of geographical and temporal scope against which incidents are considered . In addition to the number of security incidents and casualties (including death, injuries and other threats to the person), 'the general security environment in the country, population displacement and the impact of the violence on the overall humanitarian situation' should be taken into account .
There is a general consensus between the UKUT, the French Council of State, the Dutch Council of State, the German FAC and the Slovenian Supreme Court that the level of violence has to be assessed by its quantity as well as by its quality. For German courts the assessment of the quantity of the violence is a necessary starting point for assessing its quality . The reported decisions by courts and tribunals elsewhere in Europe reveal a similar concern that assessment should consider both quantity and quality. There can be no doubt that a substantial quantity of violence is a necessity without which subsidiary protection shall not be granted. But defining the threshold of Article 15(c) is not a simple matter of analysing quantitative data.
In light of the fluid case law, it would be unwise to try and lay down a fixed list of possible indicators, but from an analysis of leading cases, including Sufi and Elmi, K.A.B . (dealing with Article 3 ECHR) and the German FAC, the Dutch Council of State, the UKUT, the French National Asylum Court, the Slovenian Supreme Court (to name but a few) and by reference to UNHCR's Eligibility Guidelines on such countries as Iraq and Somalia and Afghanistan, there are three principles that should govern assessment:
a) First, the approach must be holistic and inclusive. Courts and tribunals must take into account a wide range of relevant variables.
b) Second, courts and tribunals should not limit themselves to a purely quantitative analysis of figures of civilian death and injuries etc. The approach must be qualitative as well as quantitative. When assessing quantity and quality, courts and tribunals should bear in mind the likelihood of unreported incidents and other uncertainties.
c) Third, building on the case law, which in turn absorbs insights from academic studies, courts and tribunals should look in particular to see what the evidence tells us about the indicators of situations of violence and conflict (the following is intended as a non-exhaustive list):
• The ECHR 'Sufi and Elmi Criteria':– the parties to the conflict and their relative military strengths;
– methods and tactics of warfare applied (risk of civilian casualties);
– type of weapons used;
– the geographical scope of the fighting (localised or widespread);
– the number of civilians killed, injured and displaced as a result of the fighting.
• The ability or lack of it by the State to protect its citizens against violence (where practicable, it will assist to set out the various potential actors of protection and to address their actual role)/the degree of State failure).
• Socio-economic conditions (which should include assessment of economic and other forms of assistance by international organisations and NGOs).
• Cumulative effects of long lasting armed conflicts.
In principle, these non-exhaustive indicators will apply when a general or specific risk to an applicant has to be assessed. As every single armed conflict might follow different patterns, it is of the upmost importance to remember that a list of indicators — as above — can never be exhaustive. Characteristics of an armed conflict and its civilian victims may lead to other indicators that should be taken into account.
The concept of sliding scale, derived from the Elgafaji judgment (although no specifically described as such), provides a framework for assessing the relative significance of the notions of general risk (where there is indiscriminate violence at such a high level that merely by being a civilian a person is at risk) and specific risk (where there is an individualised threat). This gives effect and context to the wording of recital (35) (ex (26)) in the preamble of the QD: the existence of a serious and individual threat to civilians generally can exceptionally be considered to be established where the degree of indiscriminate violence characterizing the armed conflict taking place reaches a high level: this is the general risk dimension of Article 15(c). If there is a general risk, the issue of credibility is not relevant; more precisely, credibility is limited to a check on whether the applicant comes from a particular country or region.
But it may still be able to succeed under Article 15(c) even when the level of indiscriminate violence is lower, if an applicant is able to show that he is specifically affected by reason of factors particular to personal circumstances: this is the specific risk dimension of Article 15(c). The sliding-scale gives shape to how specific risk is to be assessed: '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 by him to be eligible for subsidiary protection' (Elgafaji, paragraph 39, Diakité, paragraph 31). Here the assessment of credibility will be important.
Elements to be taken into account in assessing the level of indiscriminate violence have been listed above (see Section 1.3 'Indiscriminate violence').
It is clear that assessment of specific risk under Article 15(c) must proceed in a similar fashion to assessment of claims to international protection based on Article 15(a) and (b). This follows from the CJEU insistence that 'that provision [Article 15(c)] must be subject to a coherent interpretation in relation to the other two situations referred to in Article 15 of the Directive and must, therefore, be interpreted by close reference to that individualisation' . The challenge for judges in national case law to date (see Part II Section 2.31 below) is that when it comes to applying Article 15(c) to situations where the level of indiscriminate violence is not sufficiently high to place civilians in general at risk, it is often difficult to see why an applicant able to show risk-enhancing personal circumstances needs consideration under Article 15(c). As noted earlier, they may in fact be eligible for refugee protection or for subsidiary protection under Article 15(b)  or (a). It may be, therefore, that the main utility of Article 15(c) will be in cases where the issue is whether there is a general risk to all civilians.
Following Elgafaji, the French Council of State stated in Baskarathas  that it is not required that an applicant proves specific targeting because of his personal situation when the level of indiscriminate violence reaches such a degree that there are a serious and proven reasons to believe that a civilian would be at risk solely by his presence on the territory, which was according to the Court the case in Sri Lanka in Summer 2009.
The French National Asylum Court took the young age of the asylum seeker into account as an individual element in the assessment of his real risk of serious harm in several Afghan cases. According to the Court, this element is an individual risk-enhancing element for the appraisal when the level of the violence is lower. And therefore subsidiary protection was granted. The Court also took into account elements related to this young age, such as the death of the parents, lack of family links, exposure to violence and forced enlistment in one of the armed forces . Another individual element which the Court accepted as an enhanced risk arose in a case of a man from North Kivu (Democratic Republic of Congo), where the Court found that professionals who had to travel to and from Angola would face exposure to violent acts emanating from armed groups . One relevant matter here was whether the particular profession of the applicant is fundamental to his or her identity such that it would not be reasonable to expect him or her to change it in order to avoid possible harm.
The German FAC has given examples of individual circumstances which increase the threat by indiscriminate violence: e.g. if an applicant's profession forces such a person to be close to the acts of violence, such as physicians or journalists. Also personal circumstances such as religion or ethnicity can be taken into account, if they do not lead to refugee status. The FAC required also in case of such personal circumstances a high level of indiscriminate violence or a high threat to the civilian population in the area. Indicators for this can be the number of acts of indiscriminate violence, victims and severity of civilian casualties .
The High Administrative Court of Bayern did not consider the fact that the applicant belongs to the Hazara minority (Afghanistan) to be an individual 'risk-enhancing' circumstance. According to the information available to the Court, the overall situation of the Hazara, who have traditionally been discriminated against, has improved, even if traditional tensions persist and reappear from time to time. The Hazara have always lived in the provinces of Parwar and Kabul and, according to information from UNHCR, many Hazara returned to this region. Neither does an applicant's membership of the religious group of Shiites constitute an individual 'risk-enhancing' circumstance since 15 per cent of the Afghan population were Shiites .
The High Administrative Court of North Rhine-Westfalia stated that a serious and individual threat had to be met. This was only the case if general risks accumulated in such a manner that all inhabitants of a region were seriously and personally affected, or if someone was particularly affected because of individual circumstances increasing the risk. Such individual, risk-enhancing circumstances could also result from someone's membership of a group .
In HM and others, the UKUT explained its view on the reasoning of the CJEU in Elgafaji:
'The CJEU was seen to be considering in that case that a person who is at real risk of being either a specific or a more general target of indiscriminate violence may be accorded protection when the general level of violence would not be sufficient to establish the necessary risk to one who could not show any specific reason for being affected by violence unless it reached a high level .'
The Tribunal considered whether by reference to the sliding scale there could be said to be enhanced risk to civilians in Iraq who were Sunni or Shia, or Kurds, or former Baathists. It concluded that in general there could not. At paragraph 297 the Tribunal stated:
'In our judgement the other evidence relating to Sunnis and Shi'as reveals a similar picture. However, whilst for the above reasons we find the evidence as a whole insufficient to establish Sunni or Shi'a identity as in itself an 'enhanced risk category' under Article 15(c), we do accept that depending on the individual circumstances, and in particular on their facing return to an area where their Sunni or Shi'a brethren are in a minority, a person may be able to establish a real risk of Article 15(c). (They may, of course, also be able to establish a real risk of persecution under the Refugee Convention or of treatment contrary to Article 3 of the ECHR).'
The courts and tribunals having received evidence of the existence of an armed conflict in the country of origin, will need to ascertain the geographical extent of that conflict. If the indiscriminate violence throughout the country is at such a high level that persons are at Article 15(c) risk merely by virtue of being civilians, then the applicant will be entitled to subsidiary protection. However, if the area of the country affected by such a high level of indiscriminate violence is restricted in its geographical extent to part or parts of the country of origin only, then (unless the Member State in question does not apply Article 8) an applicant's ability to show real risk of serious harm in the home area under Article 15(c) merely because an applicant is a civilian will depend on whether the home area is one in which there exists such a high level of violence. The practicalities of travelling to and staying or settling in that part of the country also need to be assessed so that a finding can be made as to whether it is reasonable to expect an applicant to relocate there. Factors to take into account may include the security around the airport/town of return, together with the safety of the route that it is necessary to take to travel to the area where the conflict does not exist. In a country where internal freedom of movement is restricted, it may be that a finding will need to be made about the legality of settlement in the area. As set out earlier, if a person cannot safely reach the area of destination because of the situation of armed conflict in the country, then an Article 15(c) risk is considered as having been established in the home area.
The specific internal protection provisions of Article 8(2) refer to 'a part of the country of origin'. It goes without saying that where a finding has been made that there is a risk of serious harm due to indiscriminate violence contrary to Article 15(c), then (unless the Member State concerned does not apply Article 8) the courts and tribunals must have concluded that internal protection is unavailable.
An applicant cannot be said to have a viable internal protection alternative if the alternative part(s) of the country would either (i) also present a real risk of suffering serious harm (against which there is no effective protection); or (ii) it would be unreasonable to expect the applicant to resettle there; or (iii) the applicant could not gain practical access to such part(s) . In considering whether there exists protection against serious harm in another part(s) of the country an examination of the nature of that protection is necessary and to do this, regard must be had to the source of the protection, its effectiveness and it durability in accordance with Article 7.
Article 8(2) requires Member States to have regard to the circumstances prevailing in the country of origin at the time of taking the decision. The UKUT has found that this does not create a legal burden on the State to prove that there is a part of the country where an applicant, who has established a well-founded fear in his or her home area, could reasonably be expected to go and live. The applicant bears the legal burden but in practice, the issue of internal relocation needs to be raised by the State and it will then be for the applicant to make good an assertion that it would not be reasonable to relocate there .
There are differences between the original and recast Article 8 that have not been the subject of any examination by the CJEU so far but the changes may have practical implications. Article 8 in its original form  recognised that the threat may not exist throughout the country of origin and therefore that an applicant would not need international protection if such a person can reasonably be expected to stay in another part of the country despite technical obstacles to return. The recast QD (see above 1.8) modifies this by requiring not merely that an applicant can reasonably be expected to stay in that part of the country, but also can safely and legally travel to and gain admittance to that part of the country and can reasonably be expected to settle there. There is no longer any reference to the term 'technical obstacles' the interpretation of which had caused difficulties. There may be a strong argument for considering that the recast formulation of these aspects of the provision is meant to clarify what was implicit in the original.
The use of the word 'settle'  in the recast QD is distinct from 'stay' in the original Directive and it may be that a situation of greater stability is envisaged.
Article 8(2) in the recast QD imposes a specific duty on Member States when deciding whether an applicant has a viable internal protection alternative to obtain precise and up-to-date information from relevant sources about conditions in the proposed alternative part(s) of the country:
'[…] 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 in accordance with Article 4. To that end, Member States shall ensure that precise and up-to-date information is obtained from relevant sources, such as the United Nations High Commissioner for Refugees and the European Asylum Support Office.'
 Cf. Article 2(d) QD which requires 'well-founded fear' of persecution for the eligibility for refugee status.
 Elgafaji, op. cit., fn. 5, paragraph 37.
 Ibid., paragraphs 35 and 43. At paragraph 36 the CJEU also said that Article 15(c) has its own 'field of application' which must mean that it has additional scope to the serious harms identified in letters (a) and (b). However, with reference to Elgafaji the ECtHR indicated in the judgment of 28 June 2011, Sufi and Elmi v the United Kingdom, applications no 8319/07 and 11449/07, at paragraph 226 that it 'is not persuaded that Article 3 of the Convention, as interpreted in N.A. v UK [app.no. 25904/07, 17 July 2008] does not offer comparable protection to that afforded under the [QD]. In particular, it notes that the threshold set by both provisions may, in exceptional circumstances, be attained in consequence of a situation of general violence of such intensity that any person being returned to the region in question would be at risk simply on account of their presence there.' As it stands therefore, it is doubtful that Article 15(c) goes significantly beyond Article 3 as interpreted by the ECtHR in Sufi and Elmi.
 ECtHR, judgment of 7 July 1989, Soering v the United Kingdom, application no 14308/88, paragraph 88.
 Elgafaji, op. cit., fn. 5, paragraph 33.
 Ibid., paragraph 34.
 Cf. Article 9(2) QD which includes a non-exhaustive list of types of harm which may constitute persecution. See CJEU pending case, Case C-472/13, Andre Lawrence Shepherd v Federal Republic of Germany.
 Elgafaji, op. cit., para 28. See also CJEU pending case, Case C-562/13, Centre public d'action sociale d'Ottignies-Louvain-La-Neuve v Moussa Abdida, Advocate General Opinion delivered on 4 September 2014.
 Diakité, op. cit., fn. 7, paragraph 28.
 Ibid., paragraph 21.
 But see also Section 1.6.1 on specific risk and Section 1.6.2 on the notion of a 'sliding-scale'.
 C. Bauloz, 'The Definition of Internal Armed Conflict in Asylum Law', Journal of International Criminal Justice (2014), p.11; C. Bauloz, 'The (Mis)Use of IHL under Article 15(c) of the EU Qualification Directive', in D. Cantor and J.-F. Durieux (eds.), op. cit., p. 261.
 Court of Appeal (UK), QD (Iraq) v Secretary of State for the Home Department  EWCA Civ. 620, para 35.
 Elgafaji, op. cit., fn. 5, paragraph 34.
 Upper Tribunal, Immigration and Asylum Chamber (UK), judgment of 13 November 2012, HM and others (Article 15(c)) Iraq CG v. the Secretary of State for the Home Department,  UKUT 00409(IAC), paragraph 42.
 Bundesverwaltungsgericht (Germany), judgment of 27 April 2010, 10 C 4.09, ECLI:DE:BVerwG:2010:270410U10C4.09.0, para 34.
 Conseil d'État (France), judgment of 3 July 2009, No 320295, Office Français de Protection des Réfugiés et Apatrides c M. Baskarathas, No 320295.
 CNDA (France), judgment of 11 January 2012, M. Samadi No 11011903 C.
 CNDA (France), judgment of 28 March 2013, M. Mohamed Adan No 12017575 C.
 Administrative Court of Slovenia, judgments of 25 Sept 2013, I U 498/2012-17 and 29 Jan 2014 I U 1327/2013-10.
 Supreme Court of the Republic of Slovenia, judgment of 10 April 2014, I Up 117/2014.
 UNHCR, Safe at last, fn. 2, p. 103.
 HM and Others, op. cit., fn. 26, paragraph 292.
 See H. Lambert, 'Causation in International Protection from Armed Conflict', in D. Cantor and J.-F. Durieux (eds.), op. cit., p. 65.
 Bundesverwaltungsgericht (Germany), judgment of 17 November 2011, 10 C 13.10, ECLI: DE: BVerwG: 2011: 171 111U1 0C13.10.0, paragraph 23.
 Bundesverwaltungsgericht (Germany), judgment of 24 June 2008, 10 C 43.07, ECLI: DE: BVerwG: 2008: 240608U10C43.0 7.0, paragraph 35.
 HM and Others, op. cit., fn. 26, paragraphs 79-80.
 Ibid., paragraph 45.
 Baskarathas, op. cit., fn. 29.
 Raad van State (the Netherlands), judgment of 7 July 2008, 200802709/1, ECLI:NL:RVS:2008:BD7524.
 UNHCR, Safe at last, fn. 2, pp. 60 and 103.
 C. Bauloz, op. cit., fn. 23, p. 253 — 'Subsidiary protection under 15(c) is carefully limited ratione personae to civilian third-country nationals or civilian stateless persons not qualifying as refugees'.
 There is no fixed IHL definition but that by G. Mettraux, International Crimes and the ad hoc Tribunals (OUP, 2005) is widely seen to capture the customary law definition; it defines civilians as 'those who are not, or no longer, members of the fighting forces or of an organised military group belonging to a party to the conflict'. In IHL there is a presumption in favour of protection and in Article 50(1) of Additional Protocol I it is stated that '[i]n case of doubt whether a person is a civilian, that person shall be considered to be a civilian'. See further E. Wilmshurst and S. Breau, Perspective on the ICRC Study on Customary International Humanitarian Law (CUP, 2007), pp. 10-11, 111-112, 406.
 Albeit it is crucial to the IHL principle of distinction: the ICRC study of customary IHL states at Rule 1: 'The parties to the conflict must at all times distinguish between civilians and combatants' [J. Henckaerts and L. Doswald-Beck, Customary International Humanitarian Law (CUP, 2005)].
 Recital  QD states: 'It is necessary to introduce common criteria on the basis of which applicants for international protection are to be recognised as eligible for subsidiary protection. Those criteria should be drawn from international obligations under human rights instruments and practices existing in Member States.' Advocate General Mengozzi stated in Diakité that it is clear from the travaux preparatoires that 'the notion of subsidiary protection is derived from the international instruments concerned with human rights'.
 In its judgment of 1 July 1997, Kalac v Turkey, application No 20704/92, the ECtHR stated that 'in choosing a military career Mr Kalac was accepting of his own accord a system of military discipline that by its very nature implied the possibility of placing on him certain of the rights and freedoms of members of the armed forces limitations incapable of being imposed on civilians'; see also judgment of the ECtHR of 8 June 1976, Engel and others v the Netherlands, applications No 5100/71 and others, para 57. More generally IHRL increasingly considers that in relation to situations of armed conflict, IHL plays a complementary role and is in fact lex specialis: see Orna Ben-Naftali (ed.) International Humanitarian Law and International Human Rights Law, OUP, 2011, pp. 3-10.
 ICTY, Appeals Chamber, judgment of 29 July 2004, Prosecutor v Blaskic, Case No IT-95-14-A.
 HM and others, op. cit., fn. 26 cited also in judgment ZQ (serving soldier), op.cit. fn. 49.
 C. Bauloz, op. cit., fn. 23, argues that 'a factual definition should be preferred over fixed legal categories focusing on too rigid statuses'.
 See e.g. Commission of Human Rights, decision of 10 July 1984, Stewart v UK, application No 10044/82, paragraph 15, 'the concept that everyone's right to life shall be protected by law ' enjoins the State not only to refrain from taking life 'intentionally' but, further, to take appropriate steps to safeguard life. This case was concerned with the application of Article 2(2) of the ECHR.
 Conseil du contentieux des étrangers/Raad voor Vreemdelingenbetwistingen (Belgium), judgment of 4 December 2007, case 4460.
 CNDA (France), judgment of 24 January 2013, M. Miakhail No 12018368 C+.
 UNHCR, Statement on Subsidiary Protection Under the EC Qualification Directive for People Threatened by Indiscriminate Violence, January 2008, p. 7. Available at: http://www.refworld.org/docid/479df7472.html.
 A distinction needs to be drawn between persons recruited according the law of the country of origin (which may make military service compulsory) and persons forced to join an armed group involuntarily: see further UNHCR Guidelines on International Protection No 10: Claims to Refugee Status related to Military Service within the context of Article 1A(2) of the 1951 Convention and/or the 1967 Protocol relating to the Status of Refugees, 3 December 2013, especially paragraphs 35-41.
 E. Tsourdi, 'What Protection for Persons Fleeing Indiscriminate Violence? The Impact of the European Courts on the EU Subsidiary Protection Regime', in D. Cantor and J-F Durieux (eds), op.cit., p. 277.
 E. Tsourdi, op. cit., fn. 59, p. 281.
 E. Tsourdi, op. cit., p. 288.
 Elgafaji, op. cit., fn. 5, paragraph 28.
 Ibid, para 33.
 UNHCR, Safe at Last, fn. 2, p. 60.
 HM and others, op. cit., fn. 26, paragraph 114.
 Administrative Court of Slovenia, judgments of 25 Sept 2013, I U 498/2012-17 and 29 Jan 2014 I U 1327/2013-10.
 'The added value of Article 15(c) is its ability to provide protection from serious risks which are situational, rather than individually targeted.' UNHCR Statement on Subsidiary Protection, op. cit., fn. 57.
 Bundesverwaltungsgericht (Germany), judgment of 31 January 2013, 10 C 15.12, para 14.
 Sufi and Elmi, op. cit., fn. 14, paragraphs 210, 265-292.
 M. Mohamad Adan, op. cit, fn. 31.
 Bundesverwaltungsgericht (Germany), op. cit., para 13f; M. Mohamad Adan, op. cit.
 HM and others, op. cit., fn. 26.
 Article 7 QD — Actors of protection '1. Protection against persecution or serious harm can only be provided by: (a) the State; or (b) parties or organisations, including international organisations, controlling the State or a substantial part of the territory of the State; provided they are willing and able to offer protection in accordance with paragraph 2. 2. Protection against persecution or serious harm must be effective and of a non-temporary nature. Such protection is generally provided when the actors mentioned under points (a) and (b) of paragraph 1 take reasonable steps to prevent the persecution or suffering of serious harm, inter alia, by operating an effective legal system for the detection, prosecution and punishment of acts constituting persecution or serious harm, and when the applicant has access to such protection. 3. When assessing whether an international organisation controls a State or a substantial part of its territory and provides protection as described in paragraph 2, Member States shall take into account any guidance which may be provided in relevant Union acts.'
 Elgafaji, op.cit., fn. 5, paragraph 40.
 However, Article 7(1)(b) specifies that protection can only be afforded by non-state actors if they control the State or a substantial part of the territory of the State and are willing and able to furnish protection in accordance with Article 7(2) QD. See Supreme Administrative Court of the Czech Republic, decision of 27 October 2011, D.K.c.Ministry of Interior, Azs 22/2011.
 UNHCR, Safe at Last?, fn. 2.
 Elgafaji, op. cit., fn. 5, para 43.
 Diakité, op. cit., fn. 7, para 30.
 HM and others, op. cit., fn. 26, paragraph 45.
 Ibid., paragraph 43.
 Ibid., paragraphs 271-274.
 Ibid., paragraph 292.
 Upper Tribunal (UK), judgment of 18 May 2012, AK (Article 15(c)) Afghanistan CG v. the Secretary of State for the Home Department,  UKUT 00163, paragraph 207.
 Judgment 10 C 4.09, op. cit., fn. 28, paragraph 34.
 Judgment 10 C 13.10., op. cit., fn. 37, paragraph 23.
 Raad van State (Netherlands), judgment of 26 January 2010, 200905017/1/V2, ECLI:NL:RVS:2010:BL1483.
 Baskarathas, op. cit., fn. 29; see also CNDA, judgment of 18 October 2011, n 10003854.
 UNHCR, Safe at Last?, fn. 2, p. 104.
 Ibid., pp. 46-47.
 Ibid., p. 104.
 H. Lambert, 'The Next Frontier: Expanding Protection in Europe for Victims of Armed Conflict and Indiscriminate Violence', IJRL 2013, 224.
 ECtHR, judgment of 5 September 2013, K.A.B v Sweden, application No 886/11.
 Elgafaji, op.cit., fn. 5, paragraph 38.
 See the Advocate General's opinion in M'Bodj, op. cit., fn. 9 as to the scope of Article 15(b).
 Baskarathas, op. cit. fn. 29.
 CNDA (France), judgment of 21 March 2013, M. Youma Khan, No 12025577 C; CNDA, judgment of 2 July 2012, M. Ahmad Zai No 12006088 C; CNDA, judgment of 18 October 2011, M. Hosseini No 10003854 C+; CNDA, judgment of 3 June 2011 M. Khogyanai No 09001675 C; CNDA, judgment of 20 December 2010 M. Haidari No 10016190 C+; CNDA, Judgment of 1 September 2010, M. Habibi No 09016933 C+.
 CNDA, judgment of 5 September 2013, M. Muela No 13001980 C.
 Bundesverwaltungsgericht (Germany), judgment of 2 February 2013, BVerwG 10 C 23.12, paragraph 33.
 High Administrative Court Bayern (Germany), judgment of 3 February 2011, 13a B 10.30394.
 High Administrative Court North Rhine-Westphalia (Germany), judgment of 29 October 2010, 9 A 3642/06.A.
 HM and others, op. cit., fn. 26, paragraph 40.
 (i) is sometimes called the 'safety' limb; (ii) the 'reasonableness' limb and (iii) the 'access' limb.
 Upper Tribunal (UK), judgment of 25 November 2011, AMM and others (conflict; humanitarian crisis; returnees; FGM) Somalia v. Secretary of State for the Home Department, CG  UKUT 00445 (IAC). For the most recent decision on the situation in Mogadishu see the decision of the Upper Tribunal in MOJ and others (Return to Mogadishu) (Rev1) (CG)  UKUT 442 (IAC).
 Art. 8 — original (still applying to Ireland and the United Kingdom (see fn. 1)) provides: 'Internal protection 1. As part of the assessment of the application for international 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.'
 Which is also applied by the ECtHR: see e.g. Judgment of 11 January 2007, Salah Skeekh v Netherlands, app.no. 1948/04  ECHR 36, paragraph 141: 'The Court considers that as a precondition for relying on an internal flight alternative, certain guarantees have to be in place: the person to be expelled must be able to travel to the area concerned, to gain admittance and be able to settle there, failing which an issue under Article 3 may arise, the more so if in the absence of such guarantees there is a possibility of the expellee ending up in a part of the country of origin where he or she may be subjected to ill-treatment.'