Quick orient — first passages from judgment

Nonetheless the position at customary international law appears to be that the term "combatant" can be used to describe protagonists in both international and internal armed conflicts: see J Heinckaerts and L Doswald Beck, Customary International Law , Vol I, 2005, Rules 11-13.

The country guidance case of RQ (Afghan national army-Hizb-i-islami-risk) Afghanistan CG [2008] UKAIT 00013

ii). Whether an asylum claimant is a soldier rather than a civilian has a significant impact on risk assessment. Assessment of any harm a soldier if returned might face must take account of his particular circumstances, including when he is not in barracks or on active duty. However, in general, serving soldiers cannot expect to be protected against military violence.

iv) Enemy targeting of a soldier off duty or of members of his family is not necessarily contrary to IHL but may very often be contrary to IHL norms of military necessity, distinction and proportionality.

v) Insofar as the risk categories of NS (Iraq; perceived collaborator; relocation) Iraq CG [2007] UKAIT 00046 may cover persons who by virtue of their work have become members of the Multinational Forces or the Coalition Provisional Authority, application of its guidance will need to bear in mind that the state's duty to protect them will be very limited.

Auto-extracted from BAILII. Full structured brief in progress — the source links below give you the verbatim judgment in the meantime.

Fadli concerned a national of Algeria who claimed that on return he would be under a legal obligation to perform military service and if he refused to do so he would be subjected to a term of imprisonment of about 9 months. He submitted that if he did military service his life and that of his family would be at risk. In paras 10-18 Schiemann LJ, who gave the judgment of the Court, stated:

In Gedara the claimant was a national of Sri Lanka who had joined the police force in Sri Lanka and had in time become a police sergeant in the Intelligence Unit, his duties including gathering intelligence in the war against the LTTE. He had left at a time when he had been told by informers that his life was in danger as the LTTE wanted to kill him and his own employers had told him they could not protect him outside the high security zone of the police station. Newman J stated:

Miss Akinbolu's submissions on the Tribunal's question concerning the relevance to the appellant's case of Fadli and Gedara were threefold. First, she considered that in order to achieve consistency with more recent House of Lords and Court of Appeal authority, the Fadli principles had to be understood as applying only in situations where the armed conflict did not pose a real risk that a soldier would be a victim of war crimes. Building on the decision of the House of Lords in Sepet and Bulbul , the Court of Appeal in Krotov had recognised that a soldier could qualify as a refugee if able to show he would be forced by his commanders on return to participate in war crimes. It would be illogical, she said, if a soldier could qualify for refugee protection if he faced a real risk of being compelled to be a perpetrator of war crimes but not if he faced a real risk of being a victim of war crimes. The protection afforded by the Refugee Convention and the ECHR was designed principally to protect people from being victims of serious harm.

Secondly, she submitted, Fadli was in and of itself a conscription case and did not address that of a soldier who voluntarily enlists. The judgments in Fadli and Gedara themselves, at least if one followed what was said in Gedara , acknowledged that there could be a spectrum of situations which soldiers might face and whether they faced persecution or lack of protection would depend on the type of military harm they would face. The notion of a spectrum of situations was also consistent with IHL norms, which recognise that soldiers who are "hors de combat" are entitled to protection akin to that of a civilian.

Thirdly, she submitted, Fadli and Gedara were of limited assistance when dealing with a case in which the soldier had a family. Even if it could be said that a soldier "signed up" to face whatever type of armed conflict that arose, it could not be said that his family did so. Yet if, as in this case, the appellant being a target would make his family a target, that must surely make the harm he faced persecutory since on Frantisek Katrinak [2001] EWCA Civ 832 principles, a person could be regarded as being at real risk of serious harm by virtue of his family being at such risk.

So far as concerned the appellant's particular circumstances, Miss Akinbolu made the following main points: (i) that it was manifest the appellant had suffered past persecution (and past serious harm); (ii) that if returned to his home area he would still face persecution from those who had targeted him and his family before; (iii) wherever in Iraq he sought to relocate he would be known to be both a serving soldier and a Yazidi and the combination of these factors would put him at risk; (iv) it was not reasonable to expect him to relocate without reuniting with his close family members; but (v) wherever he chose to live with close family members, their living together as a family would increase the risk of harm both to them and him.

Mr Eadie, on the other hand, maintained that Fadli remained the governing higher court authority for cases involving returning soldiers and police (as well as ex-soldiers and ex-police). The principles subsequently established in cases such as Sepet and Bulbul and Krotov and BE (Iran ) [2008] EWCA Civ 540 were concerned, he submitted, with the very different situation of persons being forced to perpetrate (by commission or participation) war crimes. They said nothing to suggest, he submitted, that the exception to the general rule (that performing military service did not give rise to a risk of persecution) extended or should extend to those who faced being victims of war crimes.

Further, Mr Eadie submitted, it would undermine a basic principle of refugee law if soldiers could qualify as refugees by demonstrating that on return they would face being unprotected against a real risk of being victims of war crimes. It was a critical function of any state army to be able to respond to terrorist threats; and, if Miss Akinbolu's argument was accepted, it could lead to soldiers from conflict areas all over the world fleeing abroad and thereby undermining the ability of their home states to protect its own citizens, thereby creating unnecessary refugee flows.

He submitted that it was clear from the facts considered in both Fadli and Gedara that it made no difference whether a soldier would face military violence compatible with or contrary to IHL norms. In Fadli Schiemann LJ specifically addressed the case of an individual claiming a threat from terrorist organisations such as the GIA.

It was also clear from what was said by Schiemann LJ at paragraph 17 of Fadli that in his Lordship's view the general position did not alter just because the risks the soldier faced might include risks to his family.

As a result, submitted Mr Eadie, it was unnecessary to engage with any issues about a "spectrum of situations" or the type of military violence a soldier might face. Nor was it necessary to go on to consider separately issues relating to the sufficiency of protection against any (claimed) serious harm.

However, if the Tribunal was not with him on this basic point then, he submitted, it was important to bear in mind that a soldier could only succeed in a claim to international protection if able to show that he faced a real risk of serious harm personal to him.

On this point both Mr Eadie and Mr Wordsworth sought to identify relevant evidence appertaining to the appellant's case. In essence they submitted, it was clear both from the background country evidence and the approach taken by the Tribunal in KH (Article 15(c) Qualification Directive) (Iraq) [2008] UKAIT 00023 and more recently by the European Court of Human Rights (ECHR) in FK v Sweden (App. No. 3261/06 judgment of 20 January 2009) that the general situation in Iraq, albeit still one characterised by significant levels of violence, was not such that there was a real risk of serious harm to returnees, including soldiers, generally. (We interpolate here that although the Court of Appeal in QD (Iraq ) has now overturned KH , the Tribunal's findings of fact on the levels of violence was not criticised). And on the accepted facts in this appellant's case: (i) he would spend 80% of his time in barracks in or near Baghdad; (ii) if he returned to his home area, Mosul, either he would not face being targeted or even if he would, he could obtain adequate protection from the authorities locally; (iii) in any event he would be able to achieve safety by relocating either in Baghdad or elsewhere in central or southern Iraq, possibly also in the KRG.

As regards the time the appellant spent in barracks, it could not said, he submitted, that he would be exposed to a real risk of serious harm since there was little or no evidence to show soldiers in this situation in Iraq routinely faced military violence contrary to IHL norms.

In the light of these submissions we turn then to our evaluation. So far as propositions of law are concerned, we consider that Fadli states three things. First, that it is in the nature of soldiery that a soldier can be expected to face a heightened risk to his life. Second, that the Refugee Convention does not confer the status of refugee on a soldier who faces heightened risk to his life in the form of breaches of the laws of war whilst he is a soldier. Third, (and contrary to what Miss Akinbolu submitted was its ratio), whilst there was a "spectrum of situations" in which an Algerian soldier might find himself, none of them justifies identifying any exception to the rule that soldiers must expect to face a heightened risk to their life, irrespective of whether the risk constitutes violations of the rule of war.

So far as concerns whether Fadli could be distinguished from the current case on the facts, we see considerable force in Mr Eadie's submission that it cannot. The risk the appellant in Fadli was said to face was risk from the Groupe Islamlique Arme (GIA) and for the purpose of his appeal the Court accepted that it was a "terrorist" organisation indulging in acts of kidnap and torture. The Court also accepted that one situation (along the spectrum) which Fadli might face would be terrorist attacks on him in his private house, when he was on leave in his village to see his family (para 17). In historical terms at least, our appellant's situation, both in relation to risk from insurgents using terrorist tactics and their directing them at him and his family in his home area, is remarkably similar.

Fadli principles, as we have seen, were analysed in 2006 by Newman J in Gedara . We have already noted Miss Akinbolu's submissions regarding Gedara , which argued that this case had modified somewhat the application of Fadli principles. For Mr Eadie, Gedara [2006] EWHC 1690 represented clear evidence that Fadli continued to be the applicable law even post- Krotov .

We do not think it entirely clear that that Newman J in Gedara saw Fadli principles considered in isolation as continuing to apply to refugee-related cases concerning members of police intelligence as much as to soldiers. We would not dissent from Mr Eadie's submission that in Gedara Newman J seeks to show that principles similar to those which Schiemann LJ saw as applying to refugee claims by soldiers also applied to Article 3 ECHR claims by soldiers. In that way Gedara broadens the scope of Fadli principles so they apply to both protection regimes. However, by seeking to synthesise the principles applying to both legal frameworks he underlines that both depend on a concept of protection that is practical. In this way he reaffirms, in our view, the primacy of the approach taken by the House of Lords in Horvath to the concept of protection against serious harm under the Refugee Convention, which in turn drew on the ECtHR judgement in Osman v U.K . (1998) 29 EHRR 245 . Further, and to this extent agreeing with Miss Akinbolu, we think that in para 24 Newman J sought to refine the scope of application of Fadli principles to a certain extent, since he clearly accepts that whilst the "ambit" of the state's duty to protect soldiers is restricted, it is not without any scope whatsoever. The only purpose of the last two sentences in the above paragraph can have been to refer to the need in the context of soldier cases for the state to provide practical protection. What difference that makes, if any, to the issues we have to decide is another matter, to which we shall return later.

Whilst not seeking to argue that soldiers off-duty or on home leave cease to be soldiers, Miss Akinbolu sought to argue that the IHL treaty provisions treat them as being effectively "hors de combat" and therefore entitled to the same protection as civilians. IHL drew a distinction, she said, between the situation of soldiers actively or directly involved in hostilities and those who found themselves "hors de combat": those in the latter category could not legitimately be targeted by another party to the conflict in the same way as those in the former category. Further, it was necessary to consider where along the "spectrum of situations" any particular soldier's situation fell, and in this context it was relevant, she submitted, to consider to what extent he faces attacks that were contrary to IHL. If he faces a real risk of being a victim of military action contrary to IHL he faces a real risk of being persecuted and being exposed to ill-treatment contrary to Article 15(b) of the Qualification Directive and Article 3 of the ECHR.

Mr Eadie differed strongly. He did not accept that "hors de combat" provisions had application in this type of context. Nor did he agree that it was legitimate to look at a spectrum of situations. Although in his written submissions he had stated that he would not exclude the possibility that in extreme circumstances there might be a need for a different answer at least in relation to the position under (Article 2 of) the ECHR, he expressly disavowed that caveat in his oral submissions. The respondent's considered view was, he said, that it was an integral part of the military duties a solder was expected to perform that he could face a heightened risk to him of death or injury caused by the fighting. That remained so irrespective of whether his adversaries used methods and means consistent with or contrary to IHL norms. To hold otherwise, he contended, would undermine the underlying principle of state sovereignty, namely the right of a state to protect its own citizens against armed attacks. In the nature of modern armed conflicts, insurgents would often use illegitimate methods. Indeed in some cases, the more heinous the means and methods used by the enemy, the more incumbent it would be on the state to combat it, otherwise international terrorism would prevail.

We do not agree with Miss Akinbolu that soldiers who are off-duty or on home leave can be considered as "hors de combat" and therefore entitled to protection in the same way as civilians. Article 3 of GCIV reads:

In summarising similar provisions under Article 44 of API, Y Dinstein, The Conduct of Hostilities under the Law of International Armed Conflict , states:

In our view the language of such relevant provisions says nothings to suggest that being off duty per se could be a type of choice or force of circumstance necessitating protection appropriate to civilians.

However, we do think Miss Akinbolu is right to assert that IHL rules can apply differently depending on the particular circumstances surrounding the violence that a soldier may find himself subject to. The rights of combatants to take part directly in combat are subject to restrictions and correlative duties. Treacherous or perfidious attacks on individual enemy soldiers, for example, are prohibited (Dinstein, p.31). Dinstein gives as one example of unlawful combatancy the ambush of the car of SS General Heydrich in 1942 by members of the Free Czechoslovak army who were not wearing uniforms. In this connection we have already noted Article 44(3) of API. At p.48 Dinstein clarifies that whilst uniforms are not a necessary condition, it is incumbent on combatants, including irregular forces "to distinguish themselves from civilians in some way that makes them visibly different from civilians". Nor does it seem to us that such examples are confined to situations where combatants fail to identify themselves to the soldier being targeted. Any targeting must be conducted according to the principles of distinction and proportionality. Thus a soldier who may otherwise be a legitimate target might not be so if he is known to be guarding a school and the weapon aimed at him is a mortar. Similarly, it would be contrary to the laws of war for would–be attackers to fire rockets at him in his private home if it is known (or could be expected to be known) that he is residing there with civilians. To proceed with an attack on a soldier in such situations would violate the rule (as expressed in Article 48 AP1) that parties to a conflict "shall direct their operations only against military objectives". The ICRC Commentary on API states that:

Obviously the Commentary does not mean here to suggest that soldiers on leave become civilians, but rather to indicate that the civilian population does not lose its civilian character just because there are soldiers on leave who intermingle with them –unless the soldiers' numbers are large. From this it can be inferred that so far as the targeting of soldiers on leave is concerned, and subject to the number of such soldiers being small (see further on this, the Appeal Chamber in Blaskic , para 115), to proceed to target them in such a situation would be tantamount to targeting a civilian population. We shall have cause to return to the significance of this observation when considering the appellant's personal history.

However, even if we are right in regarding Gedara as taking a somewhat different view of protection than seems to be taken by Fadli , it is only the latter precedent which is binding. In the absence of other higher court authority modifying the effect of Fadli , the latter judgment is binding on the Tribunal. The question here, however, is whether subsequent higher court authority has modified the true legal position. The cases most in point, which can conveniently be termed "the refusal to perform military service cases", are, of course, Sepet and Bulbul , Krotov (which inter alia discusses Foughali (OOTH01653, 2 June 2000 and BE (Iran )).

It was Miss Akinbolu's submission that the cases involving a refusal to perform military service have established that when assessing an individual's claim to asylum, the Tribunal and courts may and do on occasions have regard to the broader principles of IHL. It had been accepted in Krotov that it was relevant to the issue of whether someone could make out a real risk of persecution to have regard to whether the harm they would face would be contrary to IHL norms, or, to use the precise language of Krotov , "the basic rules or norms of human conduct". Utilising the notion in Fadli of the "spectrum of situations" that a soldier could face, she submitted that where he faced being the victim of acts so severe as to constitute war crimes or systemic degradation or abuse, an individual soldier cannot be considered to have consented to violence or loss of life relating from those breaches. Thus, when, as in the appellant's case he was off duty with members of his family, the character of the risk he would face in that situation, being risk far removed from that he would normally face as a member of the armed forces, was such as to amount to persecution and ill-treatment.

Mr Eadie, by contrast, contended that this line of cases had no impact on the application of Fadli principles. He gave several reasons: that they all concerned conscripts, i.e. people facing involuntary military service; that in all of them the potential persecutor was not an opposing army or armed or terrorist group but the State itself; that all concerned the question of whether the punishment afforded to the national who refuses to perform military service would amount to persecution ("[a] prerequisite for the application of the principle is the prospect of punishment by the State of the failure to appear for military service"). Further, the exception carved out in these cases for persons at risk of being forced to participate in acts contrary to the basic rules of human conduct (i.e. the most fundamental rules of war) was solely intended to prevent them being forced to act as agents of serious harm or perpetrators . That is because of the IHL principle which prohibits soldiers from carrying out (or being complicit in) war crimes. His written submission stated on this point:

Hence there was nothing in any of these cases, he submitted, establishing a principle permitting a claim by a soldier of persecution based on being faced on return with exposure to a real risk of being a victim of acts contrary to basic IHL norms.

So that we can better analyse the issues, we need to remind ourselves what was said by the higher courts in the three main cases dealing with refusal to perform military service cases in the context of an asylum claim. In Sepet and Bulbul Lord Bingham stated at para 8 that:

At para 52 Lord Hoffman noted the guidance given in point 10 the EU Joint Position paper of 1996 [in terms very similar to those now set out in Article 9(2)(e) of the Qualification Directive] stated that punishment for conscientious objection might amount to persecution if, inter alia, the conditions under which service had to be performed "would require the applicant to commit war crimes or the like."

In Krotov , Potter LJ, having set out the core provisions of international humanitarian law, which prohibit, inter alia, actions such as genocide, deliberate killing and targeting of the civilian population, rape, torture, the execution and ill treatment of prisoners and the taking of civilian hostages, stated:

…

If a court or tribunal is satisfied (a) that the level and nature of the conflict, and the attitude of the relevant governmental authority towards it, has reached a position where combatants are or may be required on a sufficiently widespread basis to act in breach of the basic rules of human conduct generally recognised by the international community; (b) that they will be punished for refusing to do so and (c) that disapproval of such methods and fear of such punishment is the genuine reason motivating the refusal of an asylum seeker to serve in the relevant conflict, then it should find that a Convention ground has been established."

Rix LJ added at para 56 that he wished to:

BE (Iran) concerned the claim to international protection of a sapper from the Iranian army who in 1999 deserted rather than continue to lay anti-personnel mines in a populated part of Iranian Kurdistan. The key issue was whether, given the accepted fact that no state of war existed in that region at the time, the appellant's right of refusal to commit war crimes, and the entitlement to international protection which it attracts, extended to orders to commit any human rights violation of sufficient seriousness. Sedley LJ analysed the cases of Sepet and Bulbul and Krotov , noting as regards the former at para 25 that:

Sedley LJ went on to note that it was on the first limb of this formulation – a requirement to participate in atrocities or gross human rights abuses – that BE founded his case.

At para 41 Sedley LJ concluded that:

Dealing with the submissions regarding these cases, it is convenient first of all to eliminate points made by Miss Akinbolu and Mr Eadie that we consider devoid of merit. Her submission that Fadli was in and of itself a conscription case and did not address that of a soldier who voluntarily enlists was correct as far as it went, but if the appellant's situation on return would be that he was a voluntary enlistee who could leave the ISF of his own accord, then his claimed fear of being a victim of military violence whilst serving as a soldier would be avoidable (we shall return to this matter below), in which case the appellant's case would fall away. As regards Mr Eadie's submission that Sepet and Bulbul and Krotov concerned conscripts (not volunteers), that was correct but so did Fadli , so it is not a distinguishing feature. In any event (leaving aside for the moment the significance of a compulsory element to the return to life as a soldier) we do not see that their character as conscription cases is of any importance to deciding on the nature of the harm the persons concerned might face. Whether a soldier faces return as a conscript or (as in BE (Iran )) as a career soldier he is, for international law purposes, a soldier.

Mr Eadie made much of the fact that in the refusal to perform military service cases the potential persecutor was not, as in Fadli , an opposing army or insurgents but the state itself. However, even though it is true that in Fadli the appellant feared harm from non-state actor (GIA) terrorists, the appellant's claim placed focus on whether or not the Algerian state could be expected to protect him against such harm. The question of state responsibility arises in both types of case. Further, inasmuch as the appellant in this case feared being a victim of military violence at the hands of non-state actors, it is settled law that persecution (and serious harm under Article 15 of the Qualification Directive) can emanate from non-state as well as state actors and since 9 October 2006, indeed, that law is now codified in reg 3 of the Protection Regulations (implementing Article 6 of the Qualification Directive). We do accept, however, that in the present type of case, a state's duty of protection may be of a different kind than arises in the refusal to perform military service cases, since whilst in the latter it is necessarily the case that the state concerned is complicit in war crimes (by forcing soldiers to commit them or to participate in their commission), that will not necessarily be the case where the issue is whether the state concerned is responsible for failing to protect its soldiers against being victims of war crimes. We shall return to this point below.

As regards Mr Eadie's attachment of significance to the fact that unlike Fadli or Gedara or this case, the refusal to perform military service cases concerned persons facing a risk of punishment, he is correct up to a point, but plainly the only reason why the punishment was seen in the latter as capable of being persecutory was because the alternative was being required to perform military service contrary to the laws of war (punishment which is excessive or disproportionate is a separate type of persecutory act: see Article 9(2)(c)). One must not confuse cause with effect. Lord Bingham in Sepet and Bulbul , para 8 saw "refus[al] to undertake compulsory military service on the grounds that such service would or might require a soldier to commit atrocities or gross human rights abuses or participate in a conflict condemned by the international community" as a separate basis for being accorded refugee status that was distinct from "refusal to serve [that would] earn grossly excessive or disproportionate punishment". In addition, it seems to us that whether the case concerns a soldier refusing to perform military service (because to do so would entail being a perpetrator of or being complicit in war crimes) or a soldier refusing to perform military service (because to do so would entail being a victim of war crimes), an element of punishment will be involved in both situations. In both types of situation, if the home state decides to disregard the soldier's wishes and forces him to fight, that action has at least a punitive element. If, however, if allows him not to fight, and instead imposes a formal punishment, then it is the punishment which forms the basis of the fear of persecution.

We turn then to Mr Eadie's strongest submission which was that Fadli has established, without modification by subsequent law or case law, that acts of state persecution could not include failure by the state to protect its soldiers effectively against exposure to a heightened risk to his life irrespective of the nature of the military violence they might have to face.

When addressing this submission, we reiterate that we find Newman J's treatment of the issues in Gedara a helpful reminder of the value of seeking to resolve the difficulties we face in this case within the existing framework of basic refugee law principles, rather than seeking to strike out on some new path. In our judgment, there is no reason to seek to develop special principles of refugee law to deal with cases of soldiers and any reference to " Fadli principles" should be understood in this context. The approach to the meaning of persecution and protection (and of the need for protection to be practical) as set out in Horvath and reflected in Articles 9 and 7 of Qualification Directive ... regs 5 and 4 of the Protection Regulations is intended to have universal application and should be well able to accommodate such cases.

It is next necessary to remind ourselves that since Fadli (and indeed since Sepet and Bulbul and Krotov and Gedara ) the matter of what constitutes persecution and protection has come to be governed by the provisions of the Refugee Qualification Directive, as implemented in the Protection Regulations (which mirror the wording of Article 9). As already noted, Article 9 (2)(e) identifies as one type of persecutory harm, prosecution or punishment of a person refusing to perform military service entailing the commission of crimes or acts contrary to the laws of war. So despite Schiemann LJ relying in para 18 of Fadli on there being no distinction under the Refugee Convention turning on the effect of acts contrary to the laws of war, the Qualification Directive definition of such acts expressly relies upon such a distinction. So do the refusal to perform military service cases. Of course, in the Article 9(2)(e) context and in the refusal to perform military service cases the contemplated actor and victim of the harm are the other way round: (assuming a state's own soldiers to be "friends" and the other party to the conflict's soldiers to be "foes") they are not friend-to-foe but foe-to-friend. But both types of harm involve the use of illegitimate military violence and depend, therefore, on a distinction between legitimate and illegitimate types of military violence. As regards protection, it is perhaps wise, given slight differences from the wording of Article 7 of the Directive to state what is provided by reg 4 of the Protection Regulations. It states in its relevant parts:

A further important point is that under Article 9(1) of the Qualification Directive (see above, para 14) persecution has to be defined by reference to human rights criteria and the latter do not exclude recourse in interpreting the term to IHL norms when the situation concerned is one of armed conflict. We derive this proposition from the fact that the ECtHR in Strasbourg has made clear that in interpreting its own human rights guarantees regard must be had to the broader body of international law: see Bankovic v Belgium and Others [2001] 11 BHRC 435 , para 57 and the related fact that the broader body of international law includes IHL and, as stated in the respondent's written submissions, Articles 2, and 3 of the ECHR "are to be interpreted having regard to the principles of IHL". There can be no question, therefore, of the Fadli principles continuing to apply except insofar as they are consistent with the definition of persecution and protection contained in the Protection Regulations.

Fourth, as was made clear by Sen (which was cited in Gedara: see para 26 above), although the ECtHR has held that soldiers must be understood as accepting that military life modifies the extent to which they can be expected to face the risk of loss of life or injury in the service of their country, it is important to note that the Court expressed this as a modification of, not a negation or exclusion of, human rights guarantees applicable to soldiers. In Sen the Court stated, it is to be recalled, that:

Whilst the Court here emphasised the need to bear in mind the particular characteristics of military life and its effects on the situation of individual members of the armed forces, it said nothing to suggest that by enlisting (or being conscripted) soldiers must be understood to have consented to any type of treatment whatsoever to which they might be subjected in the armed forces or to have waived any specific human right. In particular there is nothing to suggest that they can be expected in all circumstances to put up with any kind of military harm, even a consistent pattern of harm contrary to the basic rules of human conduct. Nor is there anything said in any other Strasbourg cases dealing with soldiers (e.g. Kalac v Turkey (1997) 27 EHRR 552) to suggest that taking into account "the particular characteristics of military life" entails that soldiers cannot invoke human rights protection against military acts that are routinely contrary to the basic rules of human conduct. Indeed what it stated further on in the same para of Sen , namely:

re-emphasises that members of the armed forces have "limitations" on their rights and freedoms, not that they are denuded of them. And in deciding whether "limitations" are permissible, much is seen to hinge on the extent and degree of the interference in the right. In certain cases the Court has accepted, for example that military courts can try soldiers and, sometimes, deprive them of their liberty: see Engel v Netherlands (1976) 1 EHRR 647 , para 54, but not in others. In several cases, the Court has held that the mere fact of joining the armed forces in the knowledge it did not approve homosexuality, does not lead to a waiver of the right to a private life: see para 54, Smith and Grady v UK (2000) 29 EHRR 493 , Brown v UK , 8 July 2003 see further Ezeh and Connors v UK , App.Nos 39665/98, 40086/98, Grand Chamber, 9 October 2003; Findlay v UK (1997) 24 EHRR 221 .

We do not take Mr Eadie to have adopted the position that soldiers have no human rights, but for the sake of clarity we should state that it would be odd indeed if the ECHR were read as having no application to soldiers. The rights enshrined in the ECHR are guaranteed to "everyone…" (Article 1). Article 15 of the ECHR allows state parties to derogate from certain human rights in terms of "war or other public emergency threatening the life of the nation", but not from non-derogable human rights such as Articles 2 and 3.

We take Mr Eadie's underlying position to be that whilst soldiers have human rights, their position changes when it comes to the context of being required by their commanders to serve in an armed conflict and face military violence from their enemies. In that situation, he considers, they cannot benefit from human rights guarantees, even those that are nonderogable, such as Articles 2 and 3. We find that position difficult to square with Strasbourg case law for three reasons. The first concerns the point we have already noted that the ECHR guarantees the right to life as a right applying to everyone. The second is that leading cases have highlighted the fact that Article 2 enjoins that everyone's right to life shall be protected by law entails not just a negative obligation on the part of the state to refrain from taking life intentionally, but a positive obligation to take appropriate steps to safeguard life. In Osman v UK (2000) 29 EHRR 245 the Court stated that the State may be responsible for a breach of human rights to a person within its jurisdiction where "the authorities knew or ought to have known at the time of the existence of a real and immediate risk to the life of an identified individual or individuals from the criminal acts of a third party and that they failed to take measures within the scope of their powers which, judged reasonably, might have been expected to avoid that risk". Third, the Court has seen Article 2 as in principle applicable to soldiers. Thus in Yavus v Turkey , 25 May 2000 CD353 the Court held that although Turkey was not in breach of its obligation to the deceased soldier who had been shot and killed by an army firearm in his barracks by another soldier who had been convicted of wilful homicide prior to his being conscripted into the army, that was only because the applicants had failed to show the Osman test had been met. (We note that in a recent case, Omer Aydin v Turkey App.no. 34813/02, judgment of 25 November 2008 the Court at para 47 confirmed what it had said in Yavuz and also cited earlier cases to similar effect, Alvarez Ramon v Spa in App.no. 51192/99, 3 July 2001 and Kilinc & ors v Turkey App.no. 40145/98, 7 June 2005.)

Further, the Court has clearly held that non-derogable human rights are applicable in situations of international or internal armed conflict: see Muslim v Turkey [2006] 42 EHRR 16 and Isayev v Russia [2005] 41 EHRR 38 .

Turning from European case law to domestic case law, we must first of all examine to what extent the cases on refusal to perform military service have modified the application of Fadli principles. We accept that these cases are only concerned with soldiers who refuse to be involved in the commission of war crimes (or, as in BE (Iran ) atrocities or gross violations of human rights). We agree that none of these cases say anything to suggest that the principles they enunciate concerning reluctant perpetrators of war crimes can be automatically read across as applying conversely to reluctant victims of war crimes. However, we do think that they establish that when assessing what acts can be persecutory (or amount to treatment contrary to Article 3 ECHR) in a military context it is important to distinguish between acts which accord with the laws of war and those that do not. They lend support for the view (seemingly rejected in Fadli ) that the Refugee Convention can be understood as recognising a distinction between legitimate and illegitimate military violence. They also show that in relation to its soldiery a state cannot be understood as having no duties of protection of any kind: these cases identify that there is at least a duty on a state to protect its soldiery from being forced to commit war crimes. These cases also indicate that it is only when the military violence involved is marked by a consistent pattern of acts contrary to the laws of war that any question can arise of a soldier being at real risk as a result of having to engage in it.

Nor do we think, focusing still on UK cases, that it is the changes in our domestic law made as a result of the Qualification Directive and the further development of the law in the refusal to perform military service cases which alone operate to modify the application of Fadli . There is clear authority outside the context of refugee law, for the proposition that soldiers are entitled in certain situations to be protected from a real and immediate threat to their right to life, e.g. if called upon to give evidence without anonymity before an inquiry where they fear being targeted by insurgents: see e.g. R (A and others) v Lord Saville of Newdigate [2002] 1 WLR, 1249 and Re: Officer L [2007] UKHL 36 (31 July 2007). Even more apposite is the case concerning Private Smith in the Administrative Court (R (Catherine Smith) v Assistant Coroner for Oxfordshire & Secretary of State for Defence [2008] EWHC 694 (Admin) ) and in the Court of Appeal ( [2009] EWCA Civ 441 ). This concerned a deceased soldier who had served in Basra, Iraq. His next of kin alleged that his death had occurred as a result of a failure by the state or its agents to protect life. In the Administrative Court, Collins J, having decided that in this context the deceased remained within the jurisdiction of the U.K., turned to the question of whether soldiers could invoke the right to life:

In the Court of Appeal Sir Anthony Clarke MR giving the judgment of the Court, addressed two questions, the "jurisdiction question" and the Article 2 question. The Court's answer to both did not differ essentially from that given by Collins J. In relation to the Article 2 question, the Court formulated it in terms of whether the inquest into Private Smith's death had to conform with Strasbourg principles applied in custody cases so that the state was under a positive obligation to protect persons in a vulnerable position ( the Osman principle: see Osman v United Kingdom (2000) 29 EHRR 245 ) and thereby to set up a system which involves a practical and effective investigation of the facts and for the determination of civil liability : see para 90. Having noted that the ECtHR had applied custody principles to conscripts, the Master of the Rolls stated:

In an 11 June 2009 submission to us the two counsel for the respondent contended that the Article 2 issue in Smith related to the investigative obligation under Article 2, which was a different issue than the one before us and the Court's consideration "does not assist on the quite different facts and issues before the Tribunal, i.e. in the Soering context, and with respect to an Iraqi national returning to military duties in Iraq as a member of the Iraqi Army, with particular respect to alleged threats to an off duty Iraqi soldier." We fully accept that the issues are different, but what is relied on here is simply the basic proposition that ECHR principles are considered to have protective implications for persons who are soldiers. The Court of Appeal in Smith was in no doubt that this was indeed a basic ECHR principle.

As regards derogable human rights, in R.(on the application of Purja et al.) v. Ministry of Defence [2004] 1 WLR 289 the Court of Appeal has accepted that restrictions on the opportunity to live in married soldiers' quarters might infringe the right to a private life.

Thus, applying human rights criteria to the definition of persecution (as Article 9 of the Qualification Directive stipulates that we must) and giving effect to the principle of protection enunciated in Horvath [2000] UKHL 37 ) in Article 7 of the same Directive (reg 4 of the Protection Regulations), we see no reason in principle why a soldier who is able on the particular facts of his case to show that on return his home country's commanders could not protect him to the extent that would be practical (we shall come back to the importance of the practicality criterion later) against a real risk of becoming a victim of systemic violations of IHL (e.g. of capture and then, as a matter of routine, torture at the hands of the enemy) would not be entitled to invoke refugee protection (assuming a Refugee Convention ground was also made out) or subsidiary protection under Article 15(b) of the Qualification Directive or Article 3 protection, notwithstanding that he could be expected to face heightened risks to his life in the course of ordinary warfare.

However, nothing we have said so far amounts to support for the proposition that soldiers generally can succeed in refugee claims merely by pointing to the fact that the armed conflict to which they face having to return to as soldiers is one in which incidents of war crimes are occurring and are likely to continue to occur. It is of the highest importance to reiterate our earlier point that in order to show real risk in relation to such a broad category, it would be necessary to demonstrate not only that there would be an absence of an appropriate level of protection, but that there was a consistent pattern of such war crimes occurring. In the words of Potter LJ in Krotov :

In our view identical considerations apply in relation to persecution arising from the failure (if there is one) of a state to protect its soldiers against a real risk of war crimes.

However, the exception we have identified to the general rule that refusal to perform military service is not a basis for a refugee claim – that relating to serving soldiers facing a real risk of being the victims of war crimes – there is an added restriction. It relates back to our earlier discussion of the importance of the element of compulsion in a person being required to perform military service. Plainly in the refusal to perform military service cases, the issue of risk on return only arose because the appellants concerned feared return to countries where they faced compulsory military service, either in the form of conscription or fixed service contracts. Without the element of compulsion it would always have been open to them to avoid the feared harm (of having to participate in war crimes) by choosing not to serve. It seems to us that the same restriction must apply in relation to a soldier basing his claim on fear of being unprotected by his commanders against a real risk of being the victim of war crimes. If he is not facing compulsory military service, then it may be open to him to avoid the feared harm by quitting the army. We shall come back to how this criterion impacts on the general situation of the Iraqi armed forces and the appellant's particular case later on.

Up to this point we have concentrated in the main on the application of human rights guarantees to soldiers. But since we depend for our analysis (in part) on ECHR guarantees encompassing IHL norms (see above para 60), it is also necessary for us to demonstrate that IHL norms themselves recognise that soldiers are entitled to some degree of protection. Having regard to IHL norms also helps cast light on how we should approach the issue of practical protection by the state in the context of refugee claims. One obvious premise here is that, although recognised as complementary and overlapping bodies of law, human rights law and IHL have significant differences. And so far as protection of soldiers is concerned, we must straightaway acknowledge that for the most part the main body of IHL, in particular that contained in the four 1949 Geneva Conventions and the two 1977 protocols, is chiefly concerned with protection of civilians, not soldiers. Further, as Fleck,op.cit. notes at p.99 within the IHL framework being a soldier does put one outwith the general protection guaranteed to civilians and specified others:

On the other hand, this family of treaties exists to regulate all aspects of armed conflict and API, Article 1(1), for example, states in unqualified terms: "The High Contracting Parties undertake to respect and to ensure respect for this Protocol in all circumstances". And Article 3((2) states that:

The Rome Statute of the International Criminal Court, although for the most part specifying crimes directed against the civilian population, includes descriptions of certain crimes, e.g. genocide, that are in unqualified terms and include, as acts amounting to war crimes, "Committing outrages upon personal dignity, in particular humiliating and degrading treatment" (Article 8(b)(xx11). There are a significant number of provisions also which impose duties on parties to a conflict compliance with which clearly operates to protect each party's own combatants. We have already noted provisions designed to protect those who are hors de combat or who qualify as "protected persons". Complementing these API Article 10(1) prescribes that "[a]ll the wounded, sick and shipwrecked, to whichever Party they belong, shall be respected and protected." API, Article 16.3 stipulates that no person engaged in medical activities shall be compelled to give to anyone belonging either to an adverse Party, or to his own Party, except as required by the law of the latter party, any information concerning the wounded and the sick who are, or who have been, under his care. IHL imposes duties of protection, that is to say, not solely on civilians but also, to some degree at least, on the soldiers belonging to each party. In the Legality of the Threat or Use of Nuclear Weapons – Advisory Opinion [1996] ICJ 2 (8 July 1996) case at para 78, the ICJ noted that, whilst the "cardinal principle" of IHL is the protection of the civilian population and civilian objects, a "second principle" is that "it is prohibited to cause unnecessary suffering to combatants: it is accordingly prohibited to use weapons causing them such harm or uselessly aggravating their suffering. In application of that second principle, States do not have unlimited freedom of choice of means in the weapons they use". IHL rules are rules indicating "the normal conduct and behaviour expected of States" (para 82). We think that although the Court was specifically concerned only with the rule that a state cannot unnecessarily harm the other side's soldiers, its formulation of the rule in universal terms strongly implies that it applies also to a state's own combatants.

We have no doubt that both human rights law and IHL recognise that being a soldier obliges one to face a heightened risk to one's life and limb. A state's maintenance of its sovereignty and territorial integrity may depend upon its soldiers facing a heightened risk to their life and limb. The broader framework of international law within which both the ECHR and IHL operate includes the UN Charter, which enshrines the sovereignty of the state and its right not to be attacked and to defend itself as a fundamental principle: see Article 51. IHL treaties contain similar provisions, e.g. APII Article 3(1) states that "Nothing in this Protocol shall be invoked for the purpose of affecting the sovereignty of a State or the responsibility of the government, by all legitimate means, to maintain or re-establish law and order in the State or to defend the national unity and territorial integrity of the State".

It must also be accepted that under both human rights law and IHL it would be untenable to hold that soldiers can only be expected to face military violence kept entirely within the rules of war. Significantly international law only attaches criminal liability to those responsible for serious violations of the laws of war. Significantly as well, both the definition of "war crimes" and "crimes against humanity" employed for the purposes of establishing criminal liability under the Rome Statute of the International Criminal Court focus on war crimes "in particular" that are "large-scale" (see Article 8(1) dealing with war crimes") or "widespread or systematic" (see Article 7 dealing with crimes against humanity). Further, whilst IHL exists to regulate the conduct of armed conflict, it is predicated on the expectation that breaches will occur. IHL seeks to regulate armed conflict, not to erect tests that must be met before parties can commence fighting.

Summarising the combined effect of the provisions of human rights and IHL on this question, we consider that a soldier cannot be expected to be protected (if at all) against the risk of meeting with isolated or even a significant level of violations of the laws of war, but only against a real risk of being exposed to a consistent pattern of such violations. And under the Refugee Convention (and under the Qualification Directive and Article 3 of the ECHR) what has to be shown is both a real risk of such harm and a failure of state protection confined to that which is practical in the circumstances: see Horvath . Although (because it would require a factual situation of an exceptional kind) it may not be easy for a soldier to establish that on return he would face being unprotected (to the requisite level) against having to face a consistent pattern of military violence of an unlawful kind, if he can, then, in our view, his situation may give rise to a further exception (beyond the exception for those who face a systemic risk of having to participate in war crimes) to the general rule that soldiers cannot succeed in showing the harm they face is serious harm.

We note Mr Eadie's contention that to admit of the exception just identified would undermine the edifice of international protection which the Refugee Convention and the Human Rights Convention exist to maintain. He invoked the spectre of soldiers fleeing their armies in order to seek international protection, thereby weakening their own state's ability to protect its own citizens. We would observe in the first place that neither the House of Lords in Sepet and Bulbul nor the Court of Appeal in Krotov or BE (Iran ) appeared to see the exception for returning soldiers facing punishment for refusing to be forced to participate in war crimes (or as in BE (Iran ) atrocities or gross violations of human rights) as undermining the edifice of refugee protection. But in any event, we were certainly not presented with any evidence to suggest any real prospect of such a spectre materialising. What a soldier would have to prove, in order to be able to succeed in a refugee claim, would be an objective risk, not simply his own subjective beliefs about such risks: see Sepet and Bulbul , para 23 (per Lord Bingham). It may also be (although we cannot determine this in the absence of evidence) that the situations in which soldiers will be able to show a well-founded fear of being exposed by their superiors to a consistent pattern of military violence contrary to the rules of war would be very rare. Certainly we see nothing in the evidence that was before the Court of Appeal in Fadli (or Gedara ), for example, to suggest that there was a consistent pattern of use of military violence by the GIA in Algeria (or by the LTTE in Sri Lanka) that was contrary to the laws of war and so of sufficient severity at the relevant time (in Algeria and Sri Lanka respectively) to give rise to such a systemic risk. The Tribunal in KH (Iraq) and the ECtHR in FK v Sweden clearly did not find as a fact that the background evidence relating to Iraq demonstrated that the armed conflict there possessed such a pattern.

We see great force in Mr Eadie's contention that it cannot be right that the more heinous the enemy the more possible it becomes for soldiers to succeed in a refugee claim based on fear of being exposed to that enemy's war crimes. It cannot be right to maintain that a state is unable to deploy its soldiers to defend itself and its citizenry against an internal or external enemy just because its use of force violates IHL norms - a fortiori when that enemy violence exhibits a consistent pattern of violations of IHL norms. But in our view the concern underlying this contention can be met by applying basic principles of refugee law relating to protection.

That brings us to what we see as a real paradox to our own analysis, one demonstrating the clear need to have regard to the implications for cases concerned with situations of armed conflict, of the Horvath principle that a state can only be expected to provide practical protection. If it is only when there is a widespread and/or systemic risk of being the victim of war crimes that gives rise to an exception to the general rule that soldiers cannot expect state protection against military violence from insurgents, then it is only going to be in the most exceptional situation that such failure of protection can arise. But if the situation is most exceptional, then it is equally clear that the state's duty to protect is heavily attenuated, by virtue of its primary responsibility to protect its own citizenry.

It may be helpful to try and clarify why we think in such situations there is a protection paradox by considering in slightly more concrete fashion a hypothetical example. Let us suppose that a state has ordered its soldiers to go and fight in a part of its territory where they would face systemic and widespread targeting from lethal chemical, biological, nuclear weapons or other prohibited weapons. On our analysis two things follow. On the one hand we cannot see that soldiers are necessarily placed outwith the protection of the basic rules of human conduct or of non-derogable human rights. We remind ourselves that the basic rules of human conduct are based on peremptory norms comprising "elemental considerations of humanity", that are "erga omnes": see Legality of the Threat or use by a State of Nuclear Weapons in Armed Conflict , ICJ Advisory Opinion (8 July 1996.

On the other hand it is apparent from this very same international authority that in extreme situations the responsibility of the state to its citizenry at large cannot be such as to divest it of the right to self defence. We accept that the strict context of the Nuclear Weapons case was not a state sending soldiers to face nuclear weapons but a state contemplating using nuclear weapons; and the rule it lays down is applied only to a state's use of nuclear weapons against an enemy, but nevertheless it seems to us that the general principles the Court enunciated must be understood to have universal application. In holding that it could not rule out that a state might in an extreme case be entitled to use nuclear weapons, the Court stated at para 96 that:

It is true that the Court made clear at paras 95-6 that recourse to nuclear weapons would generally be contrary to IHL and that it noted at para 22 that "the nuclear-weapon States appearing before it either accepted, or did not dispute, that their independence to act was indeed restricted by the principles and rules of international law, more particularly humanitarian law…". (Indeed at para 86 it noted, inter alia, that among statements made to the Court was that by the United Kingdom declaring that: "[s]o far as the customary law of war is concerned, the United Kingdom has always accepted that the use of nuclear weapons is subject to the general principles of the jus in bello" (United Kingdom, CR 95/34, p.45))

It is also true that the Court further clarified that under the UN Charter the use of force will be unlawful except in limited circumstances and that even self-defence "would warrant only measures which are proportional to the armed attack and necessary to respond to it…" (para 41, citing Military and Paramilitary Activities against Nicaragua ( Nicaragua v USA) Merits, Judgment of 27 June 1986 [1986] ICJ Reports 14, para 176) and at para 42 added:

But of course, if the situation is an exceptional one in which the very survival of the state is at stake, then protective measures a state can take in relation to its own soldiers are very likely to be extremely minimal. If self-defence is at stake then a state has competing duties, including its primary duty to ensure protection of its civilian population and compliance with that duty may mean putting soldiers in harm's way with very little notice. If a state were to fail to put its soldiers in harm's way, that might well, depending on the circumstances, amount to an abandonment of the duty to defend the civilian population. (see APII, Article 3(1)) In such a context the principles of military necessity and proportionality may mean no more than taking ameliorative steps (if practicable) to minimise casualties, e.g. providing protective radiation clothing, etc.

We would observe that the position we adopt here is not inconsistent with Fadli or Gedara if one considers that what both judgments sought principally to reject was the notion that soldiers could expect protection (to use Newman J's words in Gedara at para 24) " so as to frustrate due protection being accorded to all those entitled to protection". Such a formulation of the position would appear to leave undecided the position where a state protecting its soldiers did not conflict with its role in protecting the citizenry. We should perhaps recall also that the position we take here, albeit different from that urged by Mr Eadie in the later part of his oral submissions, is very similar to that adopted in the respondent's written submissions and in the first part of his oral submissions, at least as regards the position in relation to Article 2 of the ECHR. At paragraph 33 of those written submissions it was stated: "Whilst it might not be impossible to envisage a case in which Article 2 might be engaged on protective grounds ... such a course would be wholly exceptional".

One of Mr Eadie's reasons for urging us to take an absolutist approach was that he feared that to allow any exception might, in logic, lead to a position where it could be argued that the UK government risked violation of IHL if it were to send soldiers into conflicts that might expose them to being victims of war crimes. We hope it is first of all clear from what we have said that at most the exception we have identified in the context of refugee claims arises only when there is not simply a risk of exposure to war crimes but to war crimes being committed on a widespread and/or systematic basis (such that there is a consistent pattern).

A statutory tribunal charged with assessing risk on return under the Refugee Convention, the Qualification Directive and the Human Rights Convention is not concerned to make judgments other than legal ones on the way foreign states conduct their affairs, whether the conduct relates to military or non-military matters. And as it happens we have no role at all in pronouncing on the conduct of military affairs by the United Kingdom government. Further, we have to consider not just events that have already happened; our principal task is to look at the much less certain dimension of future risk and to assess what we consider will be done by another state in relation to its soldiers, their deployment, the equipment they possess etc, bearing in mind all the time that armed conflicts are heavily affected by contingent circumstances and the ebb and flow of warfare. But, where we are faced with claims for refugee status brought by soldiers facing compulsory involvement in an armed conflict, we are concerned to determine such matters as whether or not the armed conflict is characterised by a consistent pattern of violations of the laws of war. As is clear from the refusal to perform military service cases, we must seek to reach objective conclusions based on international law norms. And in terms of general principles of international law, it has long been the case that all states are under obligations relating to the use of force, both outside as well as inside their own territory.

We also think our approach accords with the IHL principle of command responsibility. This principle or doctrine primarily concerns the responsibility of a commander to ensure that his soldiers do not commit war crimes. That is clear from the classic formulation given this principle in United States v Tomoyui Yamashita , US Military Tribunal, Manilla, 1945) 4 Law Reports of Trials of War Criminals 14, where it was held that:

However, obligations relating to the use of force do sometimes apply to those exposed to the use of force as much as those who are using it. It is well settled law that commanders have a duty to protect the safety of troops under their command (specifically affirmed by the UK in the UK Statement on Ratification of API 28 Hab 199, paragraph (h) reprinted in A Roberts and R Guelf (eds) Documents on the Laws of War 3 rd edition 2000, pp 510-412). It is difficult to see how, in principle, it could not give rise to a violation of Article 2 of the ECHR if, e.g. a commander proceeded with a plan to attack a non-vital enemy military installation using a large number of soldiers knowing that many will be killed in the attack. The loss of life would clearly be excessive compared with the direct military advantage. As such it might contravene the IHL principles of military necessity and proportionality. Another example, which might similarly be contrary to IHL rules, might be gross failure by a commander to provide soldiers with gas masks knowing they would be exposed to unlawful methods of combat in the form of lethal gas attacks on the battlefield. In such circumstances, a failure on the part of a commander to provide protective equipment might make him culpable for failing to protect his soldiers against war crimes.

Furthermore, the words "in principle" used here carry major limitations. As is clear from both the observations of both Collins J and Sir Anthony Clarke MR in the Smith case, whether the protection of Article 2 is capable of extending to a member of the armed forces will depend very much on the circumstances of the particular case. Even as regards matters relating to faulty equipment, much may depend on what resources the state concerned has and what are the practical exigencies of the armed conflict. The means of war employed by the parties and the theatre of operation (urban, jungle, desert etc) may have considerable impact on what is practicable. As already noted, a situation of armed conflict is likely to be characterised by flux and uncertainty, making it all the more important for judges assessing risks to returning soldiers to be cautious not to overlook the many contingencies that may obtain.

Whilst therefore we do not rule out that in very extreme cases the principle of military necessity might justify governments exposing their soldiers to fight in "dirty wars" in which they might become victims of violations of the laws of war on a widespread and /or systematic basis, this must still be subject to the principle of proportionality. Otherwise it would be effectively open to a dictator, for example, to send all his soldiers on a whim needlessly to die. Genocide would be no less genocide if it arose from the act of a dictator, not in sending his army to wipe out a race or tribe, but in selecting from amongst his own soldiers only those from a particular tribe or race (because, e.g. he regarded them as inferior and more dispensable) to fight in a war zone in which he knew nuclear or chemical weapons were being used on a widespread and systematic basis, without even seeking to provide them with any protective armour or equipment.

We doubt that the appellant is directly assisted much by the case of Frantisek Katrinak [2001] EWCA Civ 832 (or recital 27 of the Qualification Directive) since that concerned whether someone could base his or her claim to persecution on being the family member of someone at risk of persecution when both are in the United Kingdom, whereas in this case the appellant's surviving family members are in Iraq and are not appellants. More importantly, that case was concerned with two appellants who were civilians and had no armed conflict dimension.

We see force in Mr Eadie's and Mr Wordsworth's arguments that that it might be said that the wife of a soldier has consented to some extent to the exigencies of military life insofar as they affect her family life, which may sometimes mean, for example, joining her husband for a period of service abroad or living apart from him for a period of years. However, it is equally clear that for the purposes of assessing whether the targeting of a soldier's family would be compatible with IHL, it will often be the case that their position will be assimilated to that of the civilian population in which they live, as long as the number of soldiers present is not large. We recall here the excerpt from the ICRC Commentary on API which we cited earlier at para 48, stating that:

If the presence of soldiers themselves does not change the civilian character of the population, then we cannot see that the presence of their families would.

The relevance of human rights and IHL principles to the situation of soldiers claiming asylum has important implications for how decision-makers should approach assessment of risk to soldiers who are off duty or on leave. Outside very exceptional situations, Fadli principles apply, with the consequence that such persons cannot succeed merely by establishing that by virtue of being a soldier they would face a heightened risk to their life or limb. Nor can they succeed merely by establishing that they, in common with other soldiers, would sometimes face military violence contrary to the laws of war: they would have to show that such military violence formed a consistent pattern. However, all that we have been discussing so far has been at the level of principle as it affects the soldiers of a country generally. None of it prevents a person from showing that their own history and particular circumstances might place them at a greater risk than the generality of soldiers. In the language of the ECtHR in cases such as Vilvirajah v UK (1991) 14 EHRR 248 , they may still be able to succeed if they can show a real risk that is personal to them. In the case of an individual at real risk of being the victim of a war crime aimed at him personally, he would not need to show in addition that there was a consistent pattern of such war crimes happening in the country concerned.

We noted earlier Miss Akinbolu's submission that in Gedara , Newman J, whilst following Fadli , had clearly envisaged that soldiers fell within the ambit of the state's duty to protect its citizens at least to some extent. Her own submission was that whilst the state did not in general have a duty to protect soldiers against the risks of military violence, that could alter if the military violence concerned was contrary to the laws of war.

Mr Eadie's position, by contrast, was that under his proposed absolutist approach the issue of whether soldiers were entitled to state protection would never arise. His submission appeared to be (at least at certain points) that those who provide protection cannot expect at one and the same time to avail themselves of it.

We have considerable difficulty with Mr Eadie's approach. Even were we to accept that soldiers can be expected to face any type of harm, we do not see that such an acceptance entails denial of the duty of the state to afford them any protection whatsoever. As already intimated, there may be both military and civilian aspects to state protection of its soldiery. Just because they are soldiers who can be expected when commanded to go and fight cannot mean, for example, that their superiors might not be under a duty (subject to what is practical in the circumstances) to ensure they have adequate kit and weaponry for the fighting they are sent into. Despite their special role arising from their ability to bear arms in service of the state, soldiers are also part of a state's citizenry. Just because they are soldiers cannot mean, for example (at least in the context of an internal armed conflict), that when they are off duty they cannot look to local police to watch out for their safety as much as (if not sometimes more than) that of ordinary citizens.

To be fair to Mr Eadie, we do not think his overall position on protection was so absolutist or indeed could be. At para 20 of the respondent's written submissions it was stated:

This statement very much reflects what is said in the UK MoD Manual of the Law of Armed Conflic t at 15.6.1: "... in internal armed conflict, the law of the place, where the armed conflict takes place continues to apply...".

We note further that in reply to the specific question we posed prior to the second hearing, whether protection afforded to soldiers was always less than that a state affords to its civilians, the position of the respondent appeared to be that "... the protection afforded to a member of the state's armed forces may be the same as a matter of domestic law to that afforded to the civilian".

Whilst we agree that protection "may" (sometimes) be the same, we think in general terms that the protection soldiery can expect must in general be more limited by virtue of the nature of their military duties. That, it seems to us, is the essence of the principle expressed by the ECtHR in Sen (and echoed by the House of Lords in Gentle and Another [2008] UKHL 20 per Lord Hope at paras 18-19 and the Court of Appeal in Smith at para 31). A civilian can expect to be protected against all types of military harm; (see API, Article 51(1), Y. Dinstein, The Conduct of Hostilities under the Law of International Armed Conflict paragraph 27); a soldier cannot. Under the rules of war a soldier can be a legitimate target of military harm from his enemy. But much still depends on the context and it may be, for example, that sometimes soldiers will be entitled to greater protection (e.g. by reasonably expecting to have sentries posted at live-in barracks when there is a danger of enemy attacks).

However, as will be clear from our earlier analysis, we have difficulties with Miss Akinbolu's position as well. Whereas we accept her argument that the general lack of any state duty to protect its soldiers against being required to fight does not necessarily extend to extreme situations where commanders know their soldiers will be subjected to war crimes, it seems to us that, at least for the generality of its soldiery, the protective responsibilities in such extreme situations will be heavily limited. Of course, the state concerned may be a state that is acting illegitimately in its use of force and so acting contrary to the UN Charter and to peremptory norms of international law. If that is so, then its own conduct may itself evince a failure of its duties to protect its citizenry. But equally it may be that the state concerned is acting in legitimate self-defence. There may be states whose actions place them somewhere on the spectrum between these two types of situation. But since our reasoning must work so as to cover all cases, let us focus on the cases posing greatest difficulty for our analysis, those concerned with legitimate self-defence.

If a state under international law is acting legitimately in relation to the measures taken to defend itself, then it can be assumed that it is conducting its military affairs according to the principles of military necessity and proportionality. At the same time, where, it finds itself engaged with insurgents who are committing or intending to commit war crimes on a wide scale against the civilian population, then it will necessarily be concerned to defend itself and protect its citizenry. It seems to us that for the same reasons that the ICJ in the Nuclear Weapons case considered that even use of nuclear weapons could be consistent with a state's right to self-defence (if legitimately exercised), so a state's duties to protect its soldiers against war crimes being committed on a wide scale (internally or externally) would be extremely minimal and may amount, for example, to doing no more than trying to ensure they are provided with uniforms designed to protect against radiation. Put another way, practical protection in situations of such extremity may necessarily (in order to ensure self-defence) be extremely minimal.

As regards the appellant's grounds under Articles 2 and 3 of the ECHR, Miss Akinbolu submitted that the position as regards risk to soldiers would be the same as under the Refugee Convention and under Article 15 of the Refugee Qualification Directive. Mr Eadie broadly agreed albeit wishing to leave open that the reach of Article 3 protection might be wider than Refugee Convention protection. We see no reason to pursue this question further, save to mention that the authority of Bagdanavicus [2005] UKHL 38 must be applied. What obtains in regard to Article 3 protection will also obtain in regard to subsidiary (humanitarian) protection under Article 15(b) (para 339C(iii) of HC395) of the Qualification Directive.

As well as answers to questions of law, the Tribunal in its memorandum also sought help from the parties with answering questions concerning the number of members of the ISF killed or wounded in Iraq since 2003 until the present and arrangements made by the Government of Iraq (GOI) for the housing of its soldiers. Not all of this information was available and, of course, answers to these questions could at best only furnish part of the picture. We have also had regard to the following passage in the lCOIS report for January 2009 including the following:

The respondent also produced the report referred to in 11.57: that of the Brookings Institute Iraq Index: Tracking Variables for Reconstruction and Security in post-Saddam Iraq , 6 November 2008. It indicates that between June 2003 and 6 November 2008, 8,583 police and military personnel were killed in Iraq. Whilst accepting that neither this report nor any other source furnishes reliably precise figures, we think it safe to find, and it is not contested by Miss Akinbolu, that the latest figures for the second half of 2008 show significantly lower levels of such casualties than for the period 2004 - 2006. We also have regard to the assessment by the Tribunal in KH (Iraq) and more recently by the ECtHR in FH v Sweden , that in 2008 the levels of violence generally in Iraq were not sufficiently severe to establish a real risk of serious harm or treatment contrary to Article 3. It is not in dispute that they have fallen considerably since then.

We lack evidence as to what proportion of these deaths was a result of illegitimate (as opposed to legitimate) military violence. But even assuming that most of them were as a result of illegitimate violence, their number over the relevant period does not suffice in our view to show that the ISF soldiery is routinely exposed to a consistent pattern of military violence contrary to the laws of war.

For reasons set out earlier, that does not prevent, however, an individual ISF soldier from being able, in the particular circumstances of his case, to demonstrate that on return he would be exposed to a real risk of such violence. But it does mean he cannot succeed by pointing solely to the general experience of ISF soldiery.

We noted earlier that in general a person facing return as a soldier cannot pray in aid fear of exposure to illicit military violence unless he is in fact compelled to serve as a soldier. That may be because he is subject to compulsory conscription or because he has signed up to a contract of service requiring him to remain in the army. This observation has some importance in the Iraq context because of the considerable evidence indicating that for the most part enrolment in the ISF is voluntary.

According to the COIS report, January 2008, at paragraphs 11.01, with the fall of Saddam military service in Iraq ceased to be compulsory. The same report at para 11.03 records a 30 November 2006 report stating that:

The same report at paragraph 10.22 cites a UNAMI Report, dated 1 September – 31 October 2006 as stating that "... absenteeism is widespread and there are reports that in Kirkuk alone, half of the 5000 police force and 13,000 Army soldiers are not reporting to duty at any given time, and many fail to return to duty".

This report goes on at paragraph 11.04 to give the absenteeism figures as averaging 15% and notes that "there is currently no judicial punishment system within the Iraqi Army. Therefore Iraqi Army commanders have little leverage to compel their soldiers to combat, and soldiers and police can quit with impunity".

Paragraph 10.43 of this report cites a Centre for Strategic and International Studies (CSIS) report of 19 June 2006, which states that "... the Iraqi army does not require soldiers to sign contracts, so soldiers treat enlistments as temporary jobs". Under the heading "Military Service", the most recent report, the 2009 COIS report, January 2009 states:

We are cautious about seeking to infer from this information that it would be open to returnee ISF soldiers in every case to simply decide to leave the army. Whilst the evidence shows that a significant number of ISF soldiers desert and go AWOL, without punishment, there clearly is a system of discipline and regulations (see "Creation of a Code of Military Discipline for the New Iraqi Army" CPA/ORD/7, 7 August 2003/23) and it may be that opportunities to leave are dependent on the particular situation (e.g. a person not being returned to a front-line unit where there would be no practical possibility of desertion for some period at least; see also the observations of Sir Anthony Clarke MR in Smith , paras 104-5; although made in relation to British soldiers, they seem apt to describe the likely position for Iraqi soldiery as well).

(What we say here will not necessarily apply to Iraqi nationals who are members of the MNF or CPA. Although from the limited information available their contracts with these bodies do not bind them to compulsory service we cannot rule out that there may be circumstances where, on return, they might be handed over to these bodies and find they cannot easily leave immediately. )

One of the questions the Tribunal asked the parties to address was whether NS requires revision in the light of Fadli and the refusal to perform military service cases. Concerning the question of whether there were legal or factual reasons for the Tribunal to depart from the country guidance case of NS , Mr Eadie and Mr Wordsworth submitted there were not. It was doubtful that the risk categories identified in NS were intended to cover soldiers; in any event, someone working in an auxiliary capacity, as a translator for example (and as was the case in NS ), could not be considered as a member of a country's armed forces or otherwise as a combatant.

Albeit stating at the outset that she saw no reason for the Tribunal to depart from the guidance in NS , Miss Akinbolu did indicate, by implication at least, that she saw the risk categories identified therein as an illustration of her general point that the further removed from active participation in armed hostilities, the greater the protective scope afforded by IHL norms.

In the light of these submissions we think it possible to dispose of this question, relatively briefly. The relevancy of the principles we have identified in this case would only arise in respect of country guidance covering members of a county's armed forces or persons who are otherwise combatants. In this regard we must acknowledge that the italicised summary in NS poses a problem because it does not make clear whether it is referring solely to persons in a civilian capacity. It states (we give the summary in full because we shall have cause to consider one of the paragraphs other than "1" later on):

Before we can inquire any further into this question, we must return to a matter which we touched on earlier. The distinction between members of the armed forces and civilians is not entirely clear cut. On the one hand, there are several sources which appear to suggest that the essential criterion is the ability to carry arms. Thus persons who accompany the armed forces without being members thereof, such as war correspondents, supply contractors, members of labour units, or of services responsible for the welfare of soldiers are to be categorised as civilians: see APIII Article 4A(4) and D Fleck, Handbook of International Humanitarian Law , 319-320, 501. Private contractors and civilian employees of the armed forces who may not assume combatant functions are also to be classified as civilians unless they directly participate in hostilities (Fleck, 320). Medical personnel and pastors are also entitled to be treated as civilians, once again, it seems, because they are not entitled to bear arms.

On the other hand, there are some categories of person who are not entitled to carry arms who may nevertheless count (at least in certain circumstances) as members of the armed forces (and so can be legitimate targets) such as judges, government officials and blue collar workers (Fleck, p.314).

In the light of these observations we consider that there is a need to modify the personal scope of the guidance in NS . To the extent that those specified are, by virtue of the nature of their work, to be considered as civilians, the guidance in NS is unaffected. In our view the appellant in NS , who was a translator, was clearly a civilian. However, if a person's work for the MNF or the CPA is such that they are to be regarded as part of their armed forces, then there may be a need for them to be classified for IHL purposes -and for the purpose of assessing the risk they face on return of persecution or ill-treatment - as soldiers (however, that need may not arise if the evidence is that on return they can simply choose to terminate their employment or services). In consequence of our earlier clarification of relevant legal principles, such person will not be able to demonstrate a real risk of persecution unless they can show they face being inadequately protected by their home state (so far as is practical in the circumstances) against exposure to a consistent pattern of military violence contrary to the laws of war. In respect of the current situation in Iraq we find that the background evidence falls well short of demonstrating that Iraqi members of the MNF or the CPA per se face such a consistent pattern of military violence contrary to the laws of war.

We reiterate here the evidence cited earlier from the Brookings Institute report, 6 November 2008, which shows that the number of police and military personnel killed in Iraq is well down on previous levels. Additionally, the Iraq Body Count (IBC) report dated 28 December 2008 notes that even the figures for deaths of Iraqi police who "remain preferred targets" (because they are less well protected than foreign troops) were significantly lower than in 2006/2007 and were 928 in 2008 (by end of November). We know from other sources that there continues to be, albeit nowhere near on the scale reached in earlier periods of the armed conflict, a significant number of attacks directed against soldiers and police by insurgents that are contrary to the rule of war e.g. beheading of hostages. But their scale and extent is quite insufficient to amount to such a pattern.

That is not to say, of course that a national of Iraq who is a member of the MNF cannot establish a risk of persecution on the basis of his or her personal circumstances although he would have to show that he faced targeting in a way the MNF and the CPA in general do not.

The country guidance case of RQ (Afghan national army-Hizb-i-islami-risk) Afghanistan CG [2008] UKAIT 00013

Among the reasons Mr Eadie gave in support of his view that Fadli principles still apply in unqualified fashion was that they had been followed by the Tribunal in RQ (Afghan national army-Hizb-i-Islami-risk) Afghanistan CG [2008] UKAIT 00013 . Whilst it is true that the Tribunal in this case did cite Fadli (and Gedara ), it is not clear to us that they adhered strictly to Fadli principles when enunciating their country guidance, since, for example, they considered that serving soldiers were entitled, at least during home leave, to protection "to the Horvath standard". In any event, insofar as the Tribunal in RQ dealt with the relevant law, they did not have the benefit of the fuller submissions we had. Given that our case is not concerned with Afghanistan, we do not presume to go any further than saying that in applying the country guidance in RQ, it will henceforth be necessary to consider it in the light of our decision on points of law here.

(The Tribunal has recently published another country guidance case on Iraq, SR (Iraqi Arab Christian: relocation to KRG) Iraq CG [2009] UKAIT 00038 . We did not seek any submissions from the parties on this case, as we are satisfied that although it updates the evidence on the general country situation in the KRG and the rest of Iraq, it does not affect anything decided in this case).

It is one of the accepted facts in this case that the appellant is a Yazidi. We need briefly to summarise the background evidence relating to this group.

The Yazidis or Yazadi (Kurdish "Ezidi"), whose own name for themselves is Dasin, are adherents of a small Middle Eastern religion with ancient origins. They are primarily ethnic Kurds and most Yazidis live near Mosul, Iraq with smaller communities in Syria, Turkey, Iran, Georgia and Armenia. Estimates of their number vary from 100,000 to 500,000-600,000 (the January 2009 COIS report estimate at para 22.70) to 800,000. Although they believe in God/Allah as the creator, they worship Malak Ta'us, apparently a pre-Islamic peacock angel. Malak Ta'us is said to have been a fallen angel who repented and recreated the world that had been broken. He filled seven jars with his tears and used them to quench the fire in Hell. Malak Ta'us links to Mithraism and, through it, to Zoroastrianism. In Iraq particularly they have been oppressed and labelled as devil worshippers for centuries. Yazidis are said to be an exclusive grouping who do not intermarry even with other Kurds and accept no converts. Their twice-daily prayer services must not be performed in the presence of outsiders. The most important ritual is the annual six-day pilgrimage to the tomb of Sheikh Adii in Lalish. Their dead are buried in conical tombs.

The COIS report for January 2009, at para 22.74 cites a UNHCR Background Paper 2005 which states:

We also need to consider what is said about the Yazidis in Tribunal country guidance cases. Of background relevance here is LM (Educated women-Chaldo-Assyrians-risk) Iraq CG [2006] UKAIT 00060 . But the only country guidance case dealing in any detail with the Yazidi specifically is NH (Iraq – Yazidis) Iraq CG [2004] UKIAT 00306 , which involved an appeal heard in November 2004, which is now over four years ago. Its relevant passages state:

We must not forget, either, that the country guidance case of NS , to which we have already made reference - in fact concerned - an appellant who was accepted as a Yazidi. At paras 25-26 the Tribunal in that case recorded submissions made by the appellant's representative which in turn referred to various background source references and an expert report which had described being a Yazidi as a "risk-increasing factor". At para 33 the Tribunal stated:

In going on to find that the appellant in that case would be at risk on return, the Tribunal noted at para 40 that in addition to the fact that it had been accepted that he was of significant interest to an important insurgent organisation, "[t]here is the additional risk factor of him being a Yazidi and as we accept, being identifiable as such since he does not go to the mosque and does not pray five times a day and is understandably not prepared to change his religion. .."

With the benefit of hindsight it is easy to cast doubt on the viability of some of the NH findings. It appears from para 12 of NH that the Tribunal attached particular importance to an acceptance by the appellant's representative that "there is no evidence of specific problems for Yazidis in recent times". Whether or not that is reconcilable with what was noted in para 10 (the reference in The Times of 5 June 2003 to their being one of the most secretive and persecuted religions and the reference to "extremist attacks"), it is clear that it is no longer the case that evidence of specific problems is lacking. The later case of NS , as we have seen, identifies evidence of specific problems. It is also not clear to us that the Tribunal in NH fully appreciated, when it wrote that "the majority of the Yazidis live in the Kurdish controlled areas where law and order is generally better than elsewhere in Iraq", that the area where Yazidis numbers are largest, Mosul, is outside the KRG. In our view, the case of NH no longer provides a proper basis for assessing risk to Yazidis in Iraq. So far as NS is concerned, we need to bear in mind that since it was heard there have been further reported attacks on Yazidis.

Before turning to consider more recent materials regarding the Yazidis, it is also salient that we note what is said about the position of minorities and religious minorities in Iraq generally. In the United Nations Assistance Mission for Iraq (UNAMI) Human Rights Report for 1 Nov-31 December 2006 it is stated at para 54 that minorities in Iraq remain severely affected by the overall security situation, economic and social degradation and the existence of organised groups, militias and insurgents, all operating with near complete impunity. Minority groups feel marginalised from political participation and decision-making at the local and national levels. The European Council on Refugees and Exiles (ECRE) in their Guidelines on the Treatment of Iraqi Asylum Seekers and Refugees in Europe, 18 April 2007 states that:

Turning to the more recent evidence, the UNAMI Human Rights Report for 1 April-30 June 2007 reports that in Ninevah Governorate, in retaliation for the murder by Yazidis of a 17 year old girl reportedly accused of having a relationship with a Muslim man, vigilante groups carried out a number of retaliatory attacks, on 23 April executing several Yazidis who were on a bus. Several days later in Erbil, gangs threatened to harm or kill Yazidis found in the city. On or about 2 May, gunmen were said to have killed two Yazidi men in Mosul.

The 2007 MRG report (cited in the January 2009 COIS report at 22.72) records that:

The November 2007 UNHCR Eligibility Guidelines for assessing International Protection Needs for Iraqi Asylum Seekers report that the Yazidis have been targeted for their religious beliefs, as they are considered to be Infidels and have also been accused of collaborating with the Americans and for their "unislamic" behaviour. The report states that the situation of Yazidis in the Ninevah Governorate has deteriorated since the fall of the former regime, due to high levels of insurgent activities and ethnic/religious tensions. Targeted attacks against Yazidis include threats, assassinations and public defamation campaigns. As they are considered Infidels (or even apostates), Yazidis have been targeted by Sunni extremists present in Baghdad and towns like Sinjar, Mosul and Tel Afar (COIS, January 2009, 22.72). Most attacks against Yazidis are not reported in the international and national media or are portrayed as incidents involving Kurds. The report goes on to give examples during 2004-2005 of Yazidis who had been targeted for their (perceived) support for the US-led invasion and the presence of the MNF, considering them collaborators. Yazidis have also been targeted for "unislamic" behaviour such as smoking in public during Ramadan or selling alcohol. The report notes that over 200 Yazidi families had emigrated away from Mosul. It mentions that on 22 April 2007 a group of 23 Yazidi workers heading home from a Mosul textile factory were shot dead, Christians and Muslims travelling on the same bus were not harmed. This incident is also highlighted in the UNAMI report and in the USSD International Religious Freedom report 2007 as follows:

An HJT report describes an attack on 14 August 2007 on villages populated by Yazidis near Mosul in which over 250 were killed. It was described by US Major General Benjamin Mixon as "ethnic cleansing" and verging on attempted genocide. The same military man commented on the inability of the MNF to protect Yazidis. In September 2007 the Minority Rights Group in a February 2007 report noted various incidents of attacks against Yazidis since 2003, commenting that: "For ethnic reasons Yazidis are caught between Arabs and Kurds in Iraq. Many Yazidis suffered in the Anfal campaign alongside the Kurds …and were forced to define themselves as Arabs". This report also noted that "there had been more major attacks against the Yazidis and that more than 400 people have been killed and injured…"(COIS January 2009, 22.75). The COIS report also cites at 22.78 the Finnish Fact-Finding Mission for the period 23 October-3 November 2007 stating that: "Unemployment is a big problem for Yazidis particularly outside the KRG-administered region". In 2008/2009 to date, so far as we can tell, there have not been any major incidents similar to those which took place in April and August 2007.

Whilst the background evidence thus summarised is not comprehensive, it leads us to the following view. Whilst in general merely being a Yazidi whose home area is outside the KRG does not suffice currently to place a person at real risk of serious harm upon return to Iraq, it must be regarded as a very significant factor adding to risk and not simply as "an additional risk factor" (as it was viewed in the case of NS ). Certainly in our judgement an appellant who is a Yazidi whose home area is outside the KRG (and on the evidence most such persons will be from Mosul or environs) is currently likely to be able to demonstrate a real risk of persecution in that area unless there are special features of his or her situation which serve to reduce or modify that risk.

As regards, internal relocation for Yazidis, we are conscious that in relation to the KRG, the general situation appears significantly different. The sizeable Yazidi community there does not, in general, meet with persecution or ill treatment, although there have been incidents there too. However, we also bear in mind the UNHCR position that access to the KRG depends in general on having a family sponsor and the Tribunal country guidance case of NS and its statement that:

We have not had any submissions from either party asking us to depart from that guidance and on the evidence before us there is insufficient to modify it. Hence it will only be in the case of Yazidis found to have "…family, community and/or political links there".

In reaching the above conclusions, two particularly significant features of the background evidence before us are that it is considered likely there is an underreporting of incidents of attacks on Yazidis and that the mainstream Sunni and Shiia religious communities are continuing to scapegoat Yazidis as (either or both) infidels and collaborators with the Americans. They are being targeted very much because of their religion. We consider it too early to tell whether the lack of any major incidents in 2008 /2009 to date demonstrates that their position has improved, particularly bearing in mind the fact that the area (excluding the KRG) where they are found in largest numbers is currently the subject of considerable tension between the Arab and Kurdish communities in the run-up to a planned referendum.

It is also unlikely for very much the same reasons that Yazidis who can establish a real risk of persecution in their home area, will be able to relocate safely in central and southern Iraq. The only exception would be if it was reasonable to expect them to relocate to the KRG, where there are a very significant number of the Yazidis community of Iraq. However, for such relocation to be possible, it would have to be on the basis identified in the country guidance case of NS : see above, para 129 [see also para 136 on the country guidance case of SR ) .

Our general conclusions can be summarised as follows:

i) There is no reason to seek to develop special principles of refugee law to deal with cases of soldiers. The approach to the meaning of persecution and protection (and of the need for protection to be practical) as set out in Horvath [2000] UKHL 37 and in the Persons In Need of International Protection Regulations SI 2006/2525 is well able to accommodate such cases. The case of Fadli [2000] EWCA Civ 297 has to be read and applied in the light of in the subsequent decisions of the higher courts, including Sepet and Bulbul [2003] UKHL 15 and Krotov [2004] EWCA Civ 69 as well as in the light of the Protection Regulations.

iii). As a general rule fears a soldier may have about having to perform military service cannot give rise to a refugee claim. That rule, however, is subject to exception. One exception already identified by the Tribunal and higher courts concerns those who would face punishment for being forced to participate in acts contrary to international humanitarian law (IHL). A further exception may arise when serving soldiers face being exposed by their country's commanders to a consistent pattern of military violence contrary to the laws of war. However, where fighting of this kind is taking place the state's duty to protect its soldiery will be heavily attenuated, by virtue of its primary responsibility to defend itself and its citizenry and will in any event vary depending on a wide range of circumstances.

vi) NH (Iraq-Yazidis) Iraq CG [2004] UKIAT 00306 is no longer to be followed. Whilst being a Yazidi does not as such place a person at risk on return to central and southern Iraq, it is a significant risk factor and special reasons would need to exist for not finding that such a person faces a real risk of persecution or treatment contrary to Article 3 ECHR.

Miss Akinbolu contended that we should find that what the appellant has experienced in his home area of Mosul amounted to past persecution. Mr Eadie and Mr Wordsworth dispute that on the basis that at that time he was a soldier and that from Fadli it is clear that the things that a soldier could be expected to face included terrorist attacks on himself when on home leave and his family in his home area.

As noted earlier, we must take as our starting point that the appellant was at that time (2005) a soldier, not a civilian and so must consider first of all the human rights and IHL position. Doing so, we think it is clear that the targeting he and his family faced then was, in both human rights and IHL terms, unlawful. It was unlawful when he himself (along with his cousin) was individually targeted whilst driving a car, since there was no suggestion that his attackers wore uniform or had any way visibly identified themselves as combatants.

What then of the earlier attack on his family? Mr Eadie and Mr Wordsworth submitted that as tragic and despicable as that was, it must be considered in line with Fadli principles, as part of the heightened risks a soldier runs. We have already explained in general terms why an attack on a soldier's family in a civilian area is highly likely to be unlawful targeting in IHL terms and we have no doubt that the attack on the appellant's family was unlawful. There was no evidence to suggest that the insurgents wore anything to distinguish themselves from the civilian population or that such an attack observed the principles of military necessity, distinction or proportionally. Being unlawful military violence it amounted to persecution and serious harm and treatment proscribed by Article 3.

However, as Mr Wordsworth correctly observed, the appellant could only be said to have experienced past persecution (if at all) if it could not be said that there was effective state protection available against the evident targeting of him and his family. In this regard we would accept that at the relevant time the background evidence strongly indicated that the authorities in Mosul had their hands full even in dealing with protection of their own police against insurgent attacks. On the basis of our earlier analysis of the law, a state's duty to protect its soldiery is subject to important practical limits. However, there is not just a question here of whether the authorities were generally able to afford protection to its soldiers in the area of Mosul. There was also a question of what practical protection they were able to afford to this particular appellant/soldier. As the Tribunal stated in para 45 of IM (Sufficiency of protection) Malawi [2007] UKAIT 00071 , having reviewed Horvath and other authorities:

By virtue of his status as a soldier, the appellant could not expect the authorities to protect him in advance against being targeted for military violence by the insurgents. The evidence we have of the level and nature of military violence at that point in time (2005) falls far short of demonstrating that there was a consistent pattern of violations of the laws of war. At the same time, the Iraqi authorities soon became aware that the appellant and his family had been the victim of a targeted attack that was plainly, by virtue of its gross failure to distinguish between military and civilian persons and property, an indiscriminate attack contrary to the laws of war. Once the authorities knew that, it was reasonable to expect they would at least assess whether any special protective measures should be taken in respect of the appellant. But in fact no such steps were taken, either by the appellant's commanders or by the local authorities. It was an accepted fact that after the attack on his family and then on him and his cousin in his car, the appellant went to the local police who made clear they could not protect him. Particularly given what we know of the general situation in Mosul at that time, we consider Mr Wordsworth's submission, that the appellant had not shown protection was lacking as he had not sought protection from higher-up authorities, to have a hollow ring. At that point in time the appellant was plainly entitled to conclude that practical protection did not exist for him.

It follows from our earlier analysis of the applicable law that the appellant experienced past serious harm against which the authorities were unable to protect him and accordingly he experienced past persecution.

Turning to consider the issue of current risk, we remind ourselves that Article 4(4) (paragraph 339K of the Immigration Rules) states that the appellant's persecution is to be treated as a "serious indication" of continuing risk "unless there are good reasons to consider that such persecution will not be repeated ...".

We are satisfied that for so long as he returned to active duty and/or lived in barracks he would be safe from persecution and would be adequately protected. We have already noted background evidence indicating that the levels of attacks against the security forces have decreased considerably from what they had been. But in our judgment, there are no good reasons why the appellant would now be safe in his home area. He was clearly identified by local insurgents who blamed him for bringing the American forces into northern Mosul. Even if he sought to avoid returning to his family home (where his wife, one of his sons and his brother had been killed) and went instead to the house where he was born and his parents normally lived, the two places were not far apart. In our view it is not reasonably likely that insurgents who had gone to the lengths of attacking his family home and subsequently trying to kill him elsewhere in the city would have ceased to have an adverse interest in him. They would soon become aware where he was, anywhere within the vicinity of Mosul. We have already observed that at the time the local police did not consider they could protect him. Although Mr Wordsworth may be right to point to an improved general security situation in Iraq, we note from the later COIS report for January 2009 that even after "Operation Lion's Roar" the security situation in Mosul and its environs was considered to remain "particularly challenging". At para 9.74 it is stated:

It would appear that because of the pending referendum the situation in Mosul and its environs remains insecure and we consider the evidence does not indicate that the police or security forces, who themselves remain one of the favoured targets of insurgents in that part of Iraq, would be able to do much more for this appellant than they did before, notwithstanding that his circumstances called for additional protection.

So far as concerns whether the appellant would have a viable internal relocation alternative, however, we can firstly exclude any realistic possibility that this appellant would be able to relocate to the KRG. On the accepted facts, although he had tribal connections in that region (albeit the two Quaidi tribes there were said to be in conflict), he had no close family members there. On the basis of existing Tribunal country guidance (unaffected by any further evidence before us), the KRG authorities would not consider admitting him in order to live in their area. However, other aspects of the appellant's situation are more complicated.

To begin with, we do not consider it at all likely that those who persecuted him and his family members in his home area would have the wherewithal to pursue him beyond the Mosul area. We bear in mind that much of the insurgent activity in Iraq is fragmented and uncoordinated and the lack of any evidence to suggest that those who attacked the appellant were part of a unified network operating more widely.

We next consider his position as a returning soldier. (We do so not only because that was what the Court of Appeal directed we do, but because we consider that to be the position accepted by the respondent in advance of the latest hearing). However we must do so mindful of the fact, which we draw from the background country evidence, that service in the ISF is voluntary and there is scant evidence of punishment being visited on those who choose not to serve. That being so, we consider that if on return the appellant feared being required to serve in conflict situations that would expose him to military violence contrary to the rules of war, he would have an opportunity soon enough to quit the army. However, we note that when describing his previous military service he mentioned being sent on missions that went on for some time and we must also take account of the likely context he will face at the point of return. We accept that it is reasonably likely that initially the Iraqi authorities would return him directly to his barracks, but even if then posted on front-line duty, that would be of finite duration. And in any period during which he would be continuing to serve as an on-duty soldier we do not consider that the evidence demonstrates that he would face a real risk of systemic military violence contrary to the rules of war. For reasons given earlier, we assess the evidence as falling well short of demonstrating that serving ISF soldiers face this type of risk in Iraq currently.

NA (Palestinians - risk) Iraq CG [2008] UKAIT 00046

It is necessary to amplify the significance we attach to the appellant's Yazidi origins.

Even though it goes too far to find that Yazidis whose home area is outside the KRG are in general at risk currently, we bear in mind that we did earlier find that being a Yazidi is a very significant risk factor. Against that background we consider the appellant's particular circumstances. We have already found that he is at risk in his home area and cannot relocate to anywhere else in the KRG (albeit not for reasons to do with his Yazidi background). So he will be a person having to live outside both his home area and the KRG, i.e., away from the only significant Yazidi community in central and southern Iraq (Mosul and environs). Being a Yazidi he would not participate in the normal daily customs and practices of other Iraqis, Sunni or Shia (e.g. by going to a mosque). Even assuming he would attempt to practise his Yazidi faith privately and discreetly, his lack of participation in such public activities would inevitably mean he would attract attention from neighbours in the local area. In our judgment, it would only be a matter of time before his Yazidi identity would become known wherever he went. The mere fact that he would be identified as a Yazidi would in itself give rise to some level of risk.

So far as protection is concerned, the background evidence relating to the Yazidi does not indicate that, whatever the position as regards the ability of the Iraqi authorities to protect its citizenry generally, they have been able to afford practical protection to the Yazidis against those who have targeted them.

We remind ourselves at this point that the standard of proof in asylum-related cases is that of a reasonable degree of likelihood or substantial grounds for believing. It is not the civil standard of balance of probabilities. Considering the appellant's history of past persecution and having regard to his likely circumstances in any area of relocation within Central and Southern Iraq, we consider it reasonably likely that it will become known to the local populace that he is or was a serving soldier and is a Yazidi and that these two factors, although not enough in themselves to give rise to persecution or ill-treatment, taken together will suffice to cause him to become the target of fresh targeting for serious harm against which the Iraqi authorities will not be able to protect him. Although the author(s) of the fresh persecution may be different from those who persecuted him and his family previously, they will act out of similar motives to those his original persecutors possessed. They too will be insurgents opposed to the GoI and the MNF and insurgents who target both perceived Army collaborators and members of the Yazidi faith. The real risk of persecution which he faces will be on account of his perceived political opinion and his religion.

For the above reasons we consider that the appellant qualifies as a refugee and that his appeal should be allowed on asylum and Article 3 human rights grounds. Had he failed to show that he had a Refugee Convention reason (imputed political opinion and religion) so as to qualify as a refugee, we would have found that he was eligible for humanitarian protection under para 339C(iii) of HC395, which implements Article 15(b) of the Qualification Directive.

Signed

Dr H H Storey, Senior Immigration Judge

Quick orient — first passages from judgment

The country guidance case of RQ (Afghan national army-Hizb-i-islami-risk) Afghanistan CG [2008] UKAIT 00013

Auto-extracted from BAILII. Full structured brief in progress — the source links below give you the verbatim judgment in the meantime.

In summarising similar provisions under Article 44 of API, Y Dinstein, The Conduct of Hostilities under the Law of International Armed Conflict , states:

…

Rix LJ added at para 56 that he wished to:

Sedley LJ went on to note that it was on the first limb of this formulation – a requirement to participate in atrocities or gross human rights abuses – that BE founded his case.

At para 41 Sedley LJ concluded that:

In our view identical considerations apply in relation to persecution arising from the failure (if there is one) of a state to protect its soldiers against a real risk of war crimes.

If the presence of soldiers themselves does not change the civilian character of the population, then we cannot see that the presence of their families would.

To be fair to Mr Eadie, we do not think his overall position on protection was so absolutist or indeed could be. At para 20 of the respondent's written submissions it was stated:

The country guidance case of RQ (Afghan national army-Hizb-i-islami-risk) Afghanistan CG [2008] UKAIT 00013

It is one of the accepted facts in this case that the appellant is a Yazidi. We need briefly to summarise the background evidence relating to this group.

The COIS report for January 2009, at para 22.74 cites a UNHCR Background Paper 2005 which states:

The 2007 MRG report (cited in the January 2009 COIS report at 22.72) records that:

Our general conclusions can be summarised as follows:

NA (Palestinians - risk) Iraq CG [2008] UKAIT 00046

It is necessary to amplify the significance we attach to the appellant's Yazidi origins.

Signed

Dr H H Storey, Senior Immigration Judge

ZQ (serving soldier) Iraq CG

Get the AI brief for this case

Quick orient — first passages from judgment

Full judgment transcript

Read the full judgment elsewhere

More Immigration cases

Discussion

ZQ (serving soldier) Iraq CG

Get the AI brief for this case

Quick orient — first passages from judgment

Full judgment transcript

Read the full judgment elsewhere

More Immigration cases

Discussion