McAdam, Jane --- "The Standard of Proof in Complementary Protection Cases: Comparative Approaches in North America and Europe" [2008] UNSWLRS 50

Last Updated: 12 December 2008

The Standard of Proof in Complementary Protection Cases: Comparative Approaches in North America and Europe

Jane McAdam[∗]

This paper was presented at the ‘Critical Issues in International Refugee Law’ Research Workshop at York University, Toronto, held on 1–2 May 2008.

It will comprise part of a forthcoming publication from that event.

Abstract

Though the title of this paper implies a technical and comparative legal analysis of the standard of proof in complementary protection claims vis-à-vis Convention refugee claims, this is only part of its substance. Indeed, while the standard of proof has become a central distinguishing feature in the Canadian context between attaining protection as a ‘refugee’ or as a ‘person in need of protection’, this debate has been largely absent from the EU arena. Nevertheless, high evidentiary burdens, combined with a haphazard consideration of the three possible grounds for subsidiary protection in the EU, mean that as in Canada, subsidiary protection status cannot be regarded as a residual status for people who would be Convention refugees but for the absence of a nexus with one of the five Convention grounds. Accordingly, this paper focuses on the legal impediments to obtaining subsidiary protection in the EU that have manifested themselves in the 18 months since the Qualification Directive entered into force for the EU Member States. Its particular issue is article 15(c), which extends protection to those facing ‘a serious and individual threat to a civilian’s life or person by reason of indiscriminate violence in situations of international or internal armed conflict. This paper examines how article 15(c) has been interpreted in the jurisprudence of a number of EU Member States and demonstrates why it is not functioning effectively as a complementary form of protection.

Introduction

Though the title of this paper implies a technical and comparative legal analysis of the standard of proof in complementary protection claims vis-à-vis Convention refugee claims, this is only part of its substance. Indeed, while the standard of proof has become a central distinguishing feature in the Canadian context between attaining protection as a ‘refugee’ or as a ‘person in need of protection’,[1] this debate has been largely absent from the EU arena. Nevertheless, high evidentiary burdens, combined with a haphazard consideration of the three possible grounds for subsidiary protection in the EU, mean that as in Canada, subsidiary protection status cannot be regarded as a residual status for people who would be Convention refugees but for the absence of a nexus with one of the five Convention grounds. Accordingly, this paper focuses on the legal impediments to obtaining subsidiary protection in the EU that have manifested themselves in the 18 months since the Qualification Directive entered into force for the EU Member States.[2] Its particular issue is article 15(c), which extends protection to those facing ‘a serious and individual threat to a civilian’s life or person by reason of indiscriminate violence in situations of international or internal armed conflict.’ This provision is poorly understood, inconsistently applied across the Member States, and in some jurisdictions is the only subsidiary protection category given full consideration when a Convention claim fails.[3] Its individualized and particularized requirements impose higher evidentiary burdens on claimants than are needed to satisfy the Convention definition, and even when protection is granted, it is of a lesser quality and duration than Convention refugee status.[4] This paper examines how article 15(c) has been interpreted in the jurisprudence of a number of Member States and demonstrates why it is not functioning as a complementary form of protection.

Background

On 29 April 2004, the Member States of the European Union adopted the Directive on Minimum Standards for the Qualification and Status of Third Country Nationals or Stateless Persons as Refugees or as Persons Who Otherwise Need International Protection and the Content of the Protection Granted. This Directive represented the fourth building block in the first phase of the Common European Asylum System,[5] intended to harmonize and streamline legal standards relating to asylum in the Member States of the EU.[6]

Described as ‘unquestionably the most important instrument in the new legal order in European asylum because it goes to the heart of the 1951 Convention Relating to the Status of Refugees’,[7] the Directive sought to clarify the constitutive elements of the Convention refugee definition and the rights flowing from refugee status, and establish a harmonized approach towards people with an international protection need falling outside the scope of the Refugee Convention, known as ‘beneficiaries of subsidiary protection’. The Directive therefore has two components: clarifying the eligibility criteria for international protection, and setting out the resultant status for those who qualify. It is the first supranational instrument to elaborate a separate and distinct status for people who are not Convention refugees but are otherwise in need of protection.[8]

It is important to recall that the Directive was based on pre-existing Member State practice, and aimed simply to harmonize existing concepts by drawing on the ‘best’ elements of the Member States’ national systems.[9] It was therefore not intended as a comprehensive overhaul of protection, but rather as a partial codification of existing State practice that sought to balance the divergent political views of the various Member States.

Each Member State was supposed to have transposed the Qualification Directive into national law by 10 October 2006,[10] although as of August 2007, 12 Member States had still not transposed it in full, four had only partially transposed it,[11] and Greece had not transposed it at all.[12] Although an absence of implementing legislation should mean that the Directive’s provision apply directly where they are clear and unconditional,[13] there remain striking inconsistencies in whether, and how, the Directive is actually being applied (not least because some key provisions are not ‘clear’).[14] This necessarily undermines the process of harmonization which the Directive was intended to bring about.

Between March and July 2007, UNHCR undertook a comprehensive study of the implementation of the Qualification Directive in five Member States—France, Germany, Greece, the Slovak Republic and Sweden—which were selected as a geographical cross-section of the EU, with a variety of legal systems and institutional frameworks, and which together received almost half of all asylum applications lodged in the EU in 2006.[15] The study’s purpose was to highlight whether the Member States were adopting a consistent approach to interpreting the Directive, whether national law and practice was consistent with international standards, and whether good practices could be identified.[16] Its release in November 2007 coincided with the one-year anniversary of the deadline for transposition and the lead-up to the Commission’s report to the European Parliament and the Council (by 10 April 2008) as to whether any amendments to the Directive were required.[17] Of particular relevance to the present paper is UNHCR’s analysis of the application of article 15 on subsidiary protection, and whether the threshold for establishing a need for subsidiary protection differs in substance to claims for Convention refugee status.

Article 15

In legal terms, the inclusion of article 15 has expanded the scope of protection formally offered throughout the EU. In practice, however, narrow interpretations and procedural flaws mean that subsidiary protection is not, on the whole, increasing the numbers of people receiving protection in the Member States.[18] In Greece, for example, accelerated asylum procedures mean that subsidiary protection is not being systematically considered for asylum applicants, by contrast to France, Germany, Sweden and the UK, where a single procedure requires decision-makers to first assess an applicant’s claim in accordance with the Convention grounds, before turning to the subsidiary protection criteria if that person is not found to be a Convention refugee.[19]

Perhaps the more fundamental problem, however, is the divergent interpretations being employed across the Member States in the assessment of refugee and subsidiary protection status. On the one hand, these interpretations reflect long-standing idiosyncracies of particular States; on the other, they highlight the problems created by poorly drafted, hastily adopted and decontextualized subsidiary protection provisions in the Qualification Directive. In the Member States’ wily attempt to confine the categories of people to whom protection should be extended, they have in fact created a new area for divergence and inconsistency which undermines the very harmonization process.[20]

As anticipated by a number of commentators prior to the Directive’s entry into force,[21] the particular problem is article 15(c). This provision extends subsidiary protection to civilians who face a ‘real risk’[22] of a ‘serious and individual threat’ to their ‘life or person by reason of indiscriminate violence in situations of international or internal armed conflict’.[23] Recital 26 provides further that: ‘Risks to which a population of a country or a section of the population is generally exposed do normally not create in themselves an individual threat which would qualify as serious harm.’

Article 15(c) is not only potentially inconsistent and rendered meaningless (especially when read in conjunction with recital 26[24]), but States’ independent analysis of its meaning, apparently without regard to the interpretations being adopted in other Member States, the jurisprudential trends in the European Court of Human Rights,[25] or the guidance of UNHCR,[26] has led to vastly different recognition rates across the EU of people fleeing violence in Iraq, Chechnya and Somalia,[27] and has created legal uncertainty about the meaning of a provision that is supposed to give rise to a uniform approach.[28]

First, article 15(c) has not been transposed in a uniform manner into the national law of the five Member States examined in the UNHCR study, and of the remaining Member States, at least Lithuanian and Belgian law contain wording which differs from article 15(c).[29] France has added a requirement that the threat to the civilian is ‘direct’, and in the Slovak Republic and in Sweden, the provision is not limited to ‘civilians’. The Swedish provision initially appears broader than article 15(c) because it extends to people fearing harm in ‘other severe conflicts’, but Sweden’s restrictive interpretation of ‘internal armed conflict’ means that ‘other severe conflicts’ is used to cover situations which, in other Member States, would be encapsulated by ‘internal armed conflict’.[30] Furthermore, Swedish law requires claimants to demonstrate ‘serious abuses’ (which could include disproportionate punishment, arbitrary incarceration, physical abuse and assault, sexual abuse, social rejection, severe harassment, etc) rather than a ‘serious threat’ to life or person.[31] German law does not transpose the reference to ‘indiscriminate violence’ (although UNHCR notes that this reflects recital 26).[32] The proposed Greek law will not restrict protection from indiscriminate violence to situations of ‘international or internal armed conflict’.[33]

Secondly, and related to the absence of a harmonized approach, the elements of article 15(c)—‘serious and individual threat’ due to ‘indiscriminate violence’ in ‘situations of international or internal armed conflict’—are creating higher evidentiary burdens for claimants compared to articles 15(a) and (b) and Convention-based claims. Even though the standard of proof—‘substantial grounds ... for believing’ that the claimant faces a ‘real risk’ of serious harm if removed[34]—is identical for articles 15(a), (b) and (c), and is supposed to be comparable to the ‘well-founded fear of persecution’ standard for Convention claims, in real terms it will generally be harder to for a claimant to establish the requisite elements of article 15(c).

This is compounded by the fact that in a number of jurisdictions, it appears that articles 15(a) and (b) are not being given any (meaningful) consideration by decision-makers and that article 15(c) may be becoming the residual category for subsidiary protection claims.[35] UNHCR has queried whether the infrequent examination of article 15(b) in Swedish case law is simply ‘a matter of expediency’, or a more fundamental problem of confusion about the distinction between ‘inhuman and degrading treatment’ and ‘serious threat to life or person’.[36] If, indeed, an absence of clear doctrinal guidance is leading decision-makers to favour article 15(c),[37] then there is a risk that applicants who would otherwise fall within article 15(b), assessed in accordance with the jurisprudence of the European Court of Human Rights which does not require singling out,[38] will have to prove their claims at a higher standard. The result is that subsidiary protection claims may not be properly assessed, and that far from being a fallback status for people with a need for international protection but who do not satisfy the Convention definition, it is a more difficult status to obtain.[39]

Seeking to clarify the relationship between articles 15(a), (b) and (c), the Netherlands Council of State recently sought a preliminary ruling from the European Court of Justice as to whether article 15(c) should be interpreted as offering protection only when article 3 of the ECHR ‘also has a bearing’, or whether it instead offers ‘supplementary or other protection’, and if the latter, then what criteria should be applied ‘for determining whether a person ... runs a real risk of serious and individual threat by reason of indiscriminate violence within the terms of Article 15(c)’.[40] In UNHCR’s view, supported by the drafting records,[41] article 15(c) provides ‘added value’ to articles 15(a) and (b) by offering protection from serious risks which are situational, rather than individually focused.[42] The use of the word ‘individual’ simply indicates that a person must face a real, rather than a remote, risk, and accordingly should ‘not lead to a higher threshold and heavier burden of proof’ being imposed.[43]

(a) The requirements of article 15(c): individual threat

As anticipated prior to the Directive’s entry into force,[44] the ‘individual’ requirement in article 15(c), read in conjunction with recital 26, is being used in some Member States to deny protection to people who are at risk of serious harm but who cannot show that they are being singled out.[45] In an article published prior to the Directive’s transposition, I argued that the language of article 15(c) suggested that a person in an area of indiscriminate violence would need to show at least that he or she were personally at risk—a very problematic requirement, since indiscriminate violence is, by definition, random and haphazard. I noted that if it were interpreted even more strictly, it might require individuals to be singled out, which would establish a higher threshold than is required for either Convention-based protection or temporary protection.

These concerns have been borne out in State practice. The Swedish Migration Board requires applicants to show that they are ‘personally at risk’ because of a ‘particular circumstance’.[46] In France, applicants have to show that a personal characteristic, such as their profession, religion or wealth, is putting them at particular risk.[47] In Germany, applicants have to show that they are at greater risk than the general population or a part thereof.[48] In the UK, the AIT recently observed that the word ‘individual’ requires the applicant to demonstrate a personal risk ‘relating to the person’s specific characteristics or profile or circumstances’, despite a previous rulings from the same body that held: ‘It would be ridiculous to suggest that if there were a real risk of serious harm to members of the civilian population in general by reason of indiscriminate violence that an individual Appellant would have to show a risk to himself over and above that general risk’.[49]

As Hathaway has observed, to demand a ‘singling out’ of an applicant ‘confuses the requirement to assess risk on the basis of the claimant’s particular circumstances with some erroneous notion that refugee status must be based on a completely personalized set of facts.’[50]

In the context of claims derived from situations of generalized oppression, therefore, the issue is not whether the claimant is more at risk than anyone else in her country, but rather whether the broadly based harassment or abuse is sufficiently serious to substantiate a claim to refugee status. If persons like the applicant may face serious harm in her country, and if that risk is grounded in their civil or political status, then in the absence of effective national protection she is properly considered to be a Convention refugee.[51]

Similarly, as Goodwin-Gill and I have argued, where large groups are seriously affected ‘by the outbreak of uncontrolled communal violence, it would appear wrong in principle to limit the concept of persecution to measures immediately identifiable as direct and individual.’[52] The US Asylum Regulations dispensed with the singling out requirement in 1990, instead requiring only that a claimant show ‘a pattern or practice ... of persecution of a group persons similarly situated to the applicant,’ and his or her ‘own inclusion in, and identification with, such group of persons such that his or her fear of persecution upon return is reasonable’.[53]

The relevant question, therefore, should be whether the applicant faces a reasonable chance of persecution or (in the case of subsidiary protection) serious harm. The over-emphasis of the word ‘individual’ in article 15(c) of the Qualification Directive places a burden on applicants which goes beyond that required under the Refugee Convention and undermines the notion of subsidiary protection as a complementary basis of protection for people who cannot meet the Convention test (because, for instance, there is no link to a Convention ground).[54] As the European Court of Human Rights has observed in the context of article 3 of the ECHR, the effect of such a stringent individual requirement ‘might render the protection offered by that provision illusory if ... the applicant were required to show the existence of further special distinguishing features’.[55] This has been echoed by UNHCR in the specific context of the Qualification Directive.[56] It is not in line with comparable jurisprudence of the European Court of Human Rights on article 3 of the ECHR, where the court has expressly stated that to demonstrate a ‘real risk’ of inhuman or degrading treatment or punishment, a person does not have to establish ‘further special distinguishing features concerning him personally in order to show that he was, and continues to be, personally at risk.’[57]

(b) International or internal armed conflict

The Qualification Directive does not define ‘international or internal armed conflict’, and there is considerable divergence in its interpretation among the Member States. The UK Asylum and Immigration Tribunal (AIT) has placed considerable reliance upon an earlier draft of the provision to conclude that the phrase has to be understood according to its meaning under international humanitarian law,[58] a view which seems to be shared by the Swedish authorities.[59] There are a number of problems with this interpretation, not least because it both complicates and narrows the scope of the provision to protect people at risk of generalized violence in situations where they would be eligible for refugee protection, but for the absence of a link to a Convention ground. Because there is no singular meaning of ‘international or internal armed conflict’ in international humanitarian law,[60] determining whether or not one exists for the purposes of a determination under article 15(c) imposes a further layer of analysis which is neither straightforward nor clear-cut. Indeed, the phrase’s imprecision means that the provision’s scope remains unclear, and a strict insistence on an international humanitarian law approach takes us no closer to a definitive ‘answer’ than had a stringent ‘armed conflict’ analysis been dispensed with from the outset.[61] It simultaneously adds a complexity to the deliberation and confines the circumstances to which article 15(c) applies.[62]

The UK approach would not necessarily solve the disparity that has emerged in State practice. In France, Germany and Sweden, differing interpretations has resulted in particular conflicts being characterized as within the scope of ‘international or internal armed conflict’ in some Member States, but not in others. For example, the French authorities regard the situation in Iraq as an ‘internal armed conflict’, while the Swedish authorities do not, and within Germany, there is inconsistency across the various state jurisdictions.[63] Whereas some German courts have stated that an armed conflict only needs to be of an unpredictable duration an intensity that threatens life or limb,[64] others have required the conflict to be comparable to a civil war.[65] The upshot of these varied views is that applicants from Iraq, Chechnya and Somalia cannot be assured a consistent assessment of their situation across the Member States.[66]

Although UNHCR has acknowledged that the interpretation of ‘international or internal armed conflict’ should be informed by international humanitarian law, it has also emphasized that ‘[i]nternational protection needs arising from indiscriminate violence are not limited to situations of declared war or internationally recognized conflicts’, and that ‘[n]o formal determination by a State or an organization regarding the existence of an “international or internal armed conflict” should be required.’[67] This is the approach taken in the regional refugee instruments which extend protection to situations of generalized violence, such as the OAU Convention’s extension of protection to people fleeing ‘external aggression, occupation, foreign domination or events seriously disturbing the public order’, and the Cartagena Declaration’s application to people whose ‘lives, security or liberty have been threatened by generalized violence, foreign aggression, internal conflicts’ and so on.[68]

Furthermore, a strict international humanitarian law interpretation is at odds with the EU Temporary Protection Directive, which does not require the existence of an ‘international or internal armed conflict’ in order to be triggered.[69] As I have argued previously, for legal and logical consistency, article 15(c) ought to protect persons fleeing individually or in small groups from situations which, in a mass influx, would result in protection. The rationale behind the Temporary Protection Directive is that the size of the influx makes it inefficient or impossible to process claims in the normal way,[70] not that the nature of the threat is unique to mass influxes. Therefore to limit subsidiary protection in this way seems both illogical and inconsistent.[71] UNHCR has accordingly recommended the deletion of ‘international or internal armed conflict’ from article 15(c).[72]

Standard of proof: article 2(e)

The standard of proof for subsidiary protection is that ‘substantial grounds have been shown for believing that the person concerned, if returned to his or her country of origin ... would face a real risk of suffering serious harm as defined in Article 15’.[73] The reference to ‘substantial grounds’ stems from the case law of the European Court of Human Rights on article 3 of the ECHR and the Torture Committee on article 3 of the CAT, and was deliberately selected in order to avoid divergence between international practice and that of the Member States themselves.[74] The Torture Committee has consistently held that ‘substantial grounds’ involve a ‘foreseeable, real and personal risk’ of torture.[75] They are to be assessed on grounds that go ‘beyond mere theory or suspicion’ or ‘a mere possibility of torture’,[76] but the threat of torture does not have to be ‘highly probable’[77] or ‘highly likely to occur’.[78]

It has been held that the ‘substantial grounds’ test is intended to replicate the ‘well-founded fear’ standard under the Refugee Convention, which generally falls somewhere lower than the ‘balance of probabilities’.[79] As the AIT stated in Kacaj:

The link with the Refugee Convention is obvious. Persecution will normally involve the violation of a person’s human rights and a finding that there is real risk of persecution would be likely to involve a finding that there is a real risk of a breach of the European Convention on Human Rights. It would therefore be strange if different standards of proof applied. ... Since the concern under each Convention is whether the risk of future ill-treatment will amount to a breach of an individual’s human rights, a difference of approach would be surprising. If an adjudicator were persuaded that there was a well-founded fear of persecution but not for a reason which engaged the protection of the Refugee Convention, he would, if Mr. Tam is right, be required to reject a human rights claim if he was not satisfied that the underlying facts had been proved beyond reasonable doubt. Apart from the undesirable result of such a difference of approach when the effect on the individual who resists return is the same and may involve inhuman treatment or torture or even death, an adjudicator and the tribunal would need to indulge in mental gymnastics. Their task is difficult enough without such refinements.[80]

In that case, the AIT rejected the government’s submission that a higher standard of proof was applicable to claims under article 3 of the EHCR on the basis that:

There is nothing in the jurisprudence of the human rights’ Court or Commission which requires us to adopt a different approach to the standard applicable to the Refugee Convention; indeed, in our view, there is every reason why the same approach should be applied. Different standards would produce confusion and be likely to result in inconsistent decisions. We therefore reject the argument of the Secretary of State on this issue.[81]

It should be noted, however, that even though the UK has interpreted the standard in this way, it does not automatically follow from the wording of the Directive itself.[82] Whereas the AIT criticized my assessment of this on the grounds that it had been resolved by Kacaj,[83] the court’s view in that case, while the preferred approach, does not necessarily flow from the wording of the Directive itself. Indeed, Battjes, Carlier, and Piotrowicz and van Eck have also observed that the ‘substantial grounds’ threshold could be interpreted as setting a higher standard of proof than ‘well-founded fear’, as is the case in Canada and the US.[84] Indeed, during the drafting, Sweden sought to replace ‘substantial grounds’ with ‘well-founded fear’ (as per an earlier draft) to ensure that the same proof entitlements were established for beneficiaries of subsidiary protection as for refugees. At the time, Germany also observed that the ‘substantial grounds’ terminology might create problems of proof assessment, although argued that these could be resolved by article 4.[85] These concerns are not merely academic. In Portugal, it is presently the case that the Aliens and Border Service requires applicants to prove ‘beyond any doubt’ that flight from a general situation of insecurity is caused by individual reasons directly linked to flight.[86]

In Canada, the expression ‘substantial grounds for believing’ has been interpreted as imposing a higher standard than ‘well-founded fear’. In 2003, the Canadian Federal Court held that ‘substantial grounds for believing’ meant that the degree of risk for complementary protection claims under section 97 of the Immigration and Refugee Protection Act was to be determined ‘on the balance of probabilities’,[87] by contrast to the ‘well-founded fear’ threshold for Convention refugee claims, meaning a ‘reasonable chance or serious possibility’.[88]

This was affirmed by the Federal Court of Appeal. First, the court observed that article 97(1)(a) uses almost identical language to article 3 of the Convention against Torture, which means that the Committee against Torture’s interpretation of article 3 is highly relevant. Accordingly, the court concluded that the relevant standard was ‘on the balance of probabilities’ or ‘more likely than not’, noting that ‘the use of the word “would” requires a showing of probability’.[89] Secondly, the court indicated that the different nature of claims under section 96 compared to section 97(1)(a), such as the issue of nexus, meant that an identical standard of proof was not necessary (even though it recognized that there was ‘no rational sense’ in adopting a higher standard for the latter). The court extended the same threshold to article 97(1)(b) in the ‘absence of some compelling reason’ to the contrary.[90]

It has been suggested that an advantage of this dual test approach is that it ‘should encourage independent and separate analyses of the three different types of claims contained in the consolidated grounds of protection.’[91] While that is certainly important, there is no compelling reason why rigorous interpretation cannot occur even if the same standard of proof is applied.

In the EU, where a single standard of proof means that the ultimate focus is supposed to be on whether a real risk of serious harm exists,[92] it is in establishing that element that the burden under article 15(c) becomes particularly high. In addition to the requirements discussed above in relation to ‘individual’ threat and the meaning of an ‘international or internal armed conflict’, the German authorities have imposed a very high threshold for risk under article 15(c): ‘certain death or severest injuries’.[93] This goes far beyond what is required by article 2(e)—‘a real risk of suffering serious harm’—and is not in line with regional[94] or international interpretations of ‘real risk’.[95] It also conflicts with the AIT’s approach, namely that ‘real risk’ simply means that the risk ‘must be more than a mere possibility’ (a standard which ‘may be a relatively low one’).[96] Thus, though the German courts have recognized that the situation in Iraq satisfies the ‘armed conflict’ criterion in article 15(c), they have held that ‘there is no extreme danger which would necessitate the granting of subsidiary protection’ on an individual basis.[97]

The result is that claims considered under article 15(c) are subjected to additional evidentiary hurdles, making it more onerous for applicants to satisfy the test for subsidiary protection vis-à-vis Convention status.

---

[∗] BA (Hons) LLB (Hons) (Sydney) DPhil (Oxon); Senior Lecturer and Director of International Law Programs, Faculty of Law, University of New South Wales, Australia; Research Associate, Refugee Studies Centre, University of Oxford.
[1] Similarly, in the United States, the standard of proof is higher than the ‘reasonable possibility’ standard in asylum claims: 8 CFR §§ 208.16(c)(2), 208.13(b)(2).
[2] Council Directive 2004/83/EC of 29 April 2004 on Minimum Standards for the Qualification and Status of Third Country Nationals or Stateless Persons as Refugees or as Persons Who Otherwise Need International Protection and the Content of the Protection Granted [2004] OJ L304/12 (‘Qualification Directive’).
[3] Indeed, the Netherlands has sought a preliminary ruling from the European Court of Justice to clarify its meaning and purpose: see OJ C 8/5 of 12 January 2008; see also Decision 200702174/1 (12 October 2007) of Dutch Council of State.
[4] See Qualification Directive, arts 19–34. The issue of status is not examined here, but has been dealt with comprehensively in J McAdam, Complementary Protection in International Refugee Law (OUP, Oxford, 2007) 90–110; GS Goodwin-Gill and J McAdam, The Refugee in International Law (3rd edn, OUP, Oxford, 2007) 330 –35.
[5] The other instruments were: Council Directive 2001/55/EC of 20 July 2001 on Minimum Standards for Giving Temporary Protection in the Event of a Mass Influx of Displaced Persons and on Measures Promoting a Balance of Efforts between Member States in Receiving Such Persons and Bearing the Consequences thereof [2001] OJ L212/12; Council Directive 2003/9/EC of 27 January 2003 laying down Minimum Standards for the Reception of Asylum Seekers [2003] OJ L31/18; Council Regulation (EC) No 343/2003 of 18 February 2003 establishing the criteria and mechanisms for determining the Member State responsible for examining an asylum application lodged in one of the Member States by a third-country national [20030 OJ l50/1; Council Directive 2005/85/EC of 1 December 2005 on Minimum Standards on Procedures in Member States for Granting and Withdrawing Refugee Status [2005] OJ L326/13. For the second phase of the CEAS, see the Hague Programme (adopted 5 November 2004).
[6] Under the Protocol on the Position of the United Kingdom and Ireland, and the Protocol on the Position of Denmark, annexed to the Treaty on European Union [2002] OJ C325/5, those countries may elect not to adopt the asylum Directives. The UK and Ireland have, however, elected to adopt the Qualification Directive.
[7] H Lambert, ‘The EU Asylum Qualification Directive, Its Impact on the Jurisprudence of the United Kingdom and International Law’ (2006) 55 ICLQ 161, 161.
[8] Note that the regional OAU Convention and the Cartagena Declaration apply Convention refugee status rather than a separate status: Organization of African Unity Convention Governing the Specific Aspects of Refugee Problems in Africa (adopted 10 September 1969, entered into force 20 June 1974) 1001 UNTS 45; Cartagena Declaration on Refugees (22 November 1984) in Annual Report of the Inter-American Commission on Human Rights OAS Doc OEA/Ser.L/V/II.66/doc.10, rev.1, 190–93 (1984–85).
[9] ‘Explanatory Memorandum’ in Commission of the European Communities Proposal for a Council Directive on Minimum Standards for the Qualification and Status of Third Country Nationals and Stateless Persons as Refugees or as Persons Who Otherwise Need International Protection COM (2001) 510 final (12 September 2001) 5. Note that it was drafted prior to 1 May 2004 enlargement of the EU and hence relied on the State practice of the 15 Member States at that time.
[10] Qualification Directive, art 38. In accordance with articles 1 and 2 of the Protocol on the Position of Denmark annexed to the Treaty on European Union [2002] OJ C325/5 and the Treaty establishing the European Community [2002] OJ C325/33, the Qualification Directive does not apply to Denmark (see Qualification Directive, recital 40).
[11] JRS (August 2007) 8.
[12] UNHCR, Asylum in the European Union: A Study of the Implementation of the Qualification Directive (UNHCR, Brussels, 2007) 9 (‘UNHCR Study’).
[13] ECJ 1970, 1213 ff, cited in UNHCR Study, 9.
[14] See OJ C 8/5 of 12 January 2008; see also Decision 200702174/1 (12 October 2007) of Dutch Council of State.
[15] UNHCR Study, 8.
[16] UNHCR Study, 8.
[17] Qualification Directive, art 37. The confidential report has been prepared and is currently before the Parliament and Council.
[18] UNHCR Study, 11. Although Sweden grants subsidiary protection to very large numbers of people, this has not greatly affected numbers overall, given that this continues Sweden’s historical practice of favouring subsidiary protection categories over Convention status, and Sweden’s recognition rate for Convention refugees remains comparatively very low: 81.
[19] UNHCR Study, page ref for Greece procedure; F/G comparison; UK.
[20] On the drafting history, see J McAdam, Complementary Protection in International Refugee Law (OUP, Oxford, 2007) ch 2.
[21] See eg J McAdam ‘The European Union Qualification Directive: The Creation of a Subsidiary Protection Regime’ (2005) 17 Intl J of Refugee Law 461; UNHCR Annotated Comments; Klug 616–19; Gil-Bazo; general discussion at the conference The Qualification Directive: Central Themes, Problem Issues, and Implementation in Selected Member States (Centre for Migration Law, Radboud University, Nijmegen, 6 October 2006); see the published papers in K Zwaan (ed), The Qualification Directive: Central Themes, Problem Issues, and Implementation in Selected Member States (Wolf Legal Publishers, Nijmegen, 2007), especially M Garlick, 62 –64. For the drafting history of the provision, see J McAdam, Complementary Protection in International Refugee Law (OUP, Oxford, 2007), ch 2.
[22] Qualification Directive, art 2(e).
[23] The standard is ‘substantial grounds for believing’: art 2(e).
[24] UNHCR has therefore recommended the deletion of recital 26: UNHCR Study, 74.
[25] eg Salah Sheekh v The Netherlands App No 1948/04 (11 January 2007).
[26] UNHCR Study, 14 (referring, for example, to UNHCR’s Annotated Comments (January 2005).
[27] For example, the percentage of Iraqi asylum claimants granted Convention refugee status at first instance in the first quarter of 2007 was as follows: 16.3% (Germany), 1.7% (Sweden), 0% (Greece, Slovak Republic). The percentage granted subsidiary protection status was: 1.1% (Germany), 73.2% (Sweden), 0% (Greece, Slovak Republic). See UNHCR Study, 13.
[28] See one suggestion for an international judicial commission: AM North and J Chia, ‘Towards Convergence in the Interpretation of the Refugee Convention: A Proposal for the Establishment of an
International Judicial Commission for Refugees’ in J McAdam (ed), Forced Migration, Human Rights and Security (Hart Publishing, Oxford 2008).
[29] Law on the Legal Status of Aliens (29 April 2004) No IX-2206 (Official Gazette No 73-2539, 3 April 2004) art 87 (Lithuania); Loi modifiant la loi du 15 décembre 1980 sur l’accès au territoire, le séjour, l’établissement et l’éloignement des étrangers (15 September 2006) art 26 (Belgium), which omits the ‘individual’ requirement.
[30] UNHCR Study, 67–68. This is despite drafting records which suggested that ‘other severe conflicts’ would extend to political instability in the home State and a consequent lack of safeguards for basic human rights, including where the State is not a party to the conflict.
[31] See UNHCR Study, 68.
[32] UNHCR Study, 68.
[33] UNHCR Study, 69, referring to the draft Presidential Decree, art 52(c).
[34] Qualification Directive, art 2(e).
[35] UNHCR Study, 12, 70 (Swedish practice). This is somewhat surprising, given the extensive (pre-existing and continuing) jurisprudence of the European Court of Human Rights on article 3 of the ECHR, which parallels article 15(b) of the Qualification Directive and which provides a rich basis for interpretation of cases under that head.
[36] UNHCR Study, 70–71.
[37] UNHCR Study, 12, 70–71.
[38] Salah Sheekh v The Netherlands App No 1948/04 (11 January 2007).
[39] There are parallels with the Canadian position.
[40] OJ C 8/5 of 12 January 2008; see also Decision 200702174/1 (12 October 2007) of Dutch Council of State.
[41] See J McAdam, Complementary Protection in International Refugee Law (OUP, Oxford, 2007) 70–78.
[42] UNHCR Statement, 5.
[43] UNHCR Statement, 6. See also the discussion of this standard of proof in Amnesty International (German section) and others, ‘Joint Opinion on the Legislation to Implement EU Directives on Residence and Asylum Law’ (August 2007) <http://www.proasyl.de/fileadmin/proasyl/fm_redakteure/Englisch/Joint_Opinion_Eu_directives.pdf> (2 April 2008).
[44] J McAdam, ‘The European Union Qualification Directive: The Creation of a Subsidiary Protection Regime’ (2005) 17 Intl J of Refugee Law 461, 481.
[45] UNHCR Study, 71ff, citing the approach of authorities in France, Germany and Sweden. The vast majority of Member States supported the requirement, on the grounds that it would avoid ‘an undesired opening of the scope of this subparagraph’: 12382/02 ASILE 47 (30 September 2002) para 4. Lithuania, Belgium and Finland do not have an ‘individual’ requirement: see respectively Law on the Legal Status of Aliens (29 April 2004) No IX-2206 (Official Gazette No 73-2539, 3 April 2004) art 87 (Lithuania); Loi modifiant la loi du 15 décembre 1980 sur l’accès au territoire, le séjour, l’établissement et l’éloignement des étrangers (15 September 2006) art 26 (Belgium); Aliens Act 2004, section 89.
[46] UNHCR Study, 72. [see specific decision there]
[47] See decisions referred to in UNHCR Study, 73.
[48] UNHCR Study, 73; recall that this is due to the way recital 26 has been combined with art 15(c) in section 60(7) of the Residence Act 2004.
[49] Lukman Hameed Mohamed v Sect’y of State for the Home Dept AA/14710/2006 (unreported, 16 August 2007) (cited in UNHCR Statement, 6).
[50] JC Hathaway, The Law of Refugee Status (Butterworths, Toronto, 1991) 91–92 (citations omitted).
[51] JC Hathaway, The Law of Refugee Status (Butterworths, Toronto, 1991) 97 (citations omitted).
[52] GS Goodwin-Gill and J McAdam, The Refugee in International Law (OUP, Oxford, 2007) 129. See the reference there in fn 364 to R v Sect’y of State for the Home Dept, ex p Jeyakumaran (No CO/290/84, QBD, unreported, 28 June 1985).
[53] 8 CFR §208.13(b)(2)(iii)—asylum (emphasis supplied); §208.16(b)(2)—withholding.
[54] UNHCR has stressed the importance of a full and inclusive interpretation of the refugee definition in the Convention, including recognizing its applicability in situations of generalized violence and armed conflict where a nexus to at least one of the five Convention grounds can be demonstrated: UNHCR Study, 99; see also UNHCR, ‘UNHCR Statement: Subsidiary Protection under the EC Qualification Directive for People Threatened by Indiscriminate Violence’ (January 2008) 5.
[55] Salah Sheekh v The Netherlands App No 1948/04 (11 January 2007) para 148.
[56] UNHCR Study, 74.
[57] Salah Sheekh v The Netherlands App No 1948/04 (11 January 2007) para 148.
[58] AIT 2008, para 60.
[59] UNHCR Study, 77.
[60] See eg International Committee of the Red Cross, ‘How is the Term “Armed Conflict” Defined in International Humanitarian Law? (Opinion Paper, March 2008).
[61] Note the reversal of the position of Hugo Storey, a judge in the AIT 2008 decision: H Storey and others, ‘Complementary Protection: Should There Be a Common Approach to Providing Protection to Persons Who Are Not Covered by the 1951 Geneva Convention?’ (Joint ILPA/IARLJ Symposium, 6 December 1999) (copy with author) 15.
[62] There must be a minimum level of intensity, ‘parties to the conflict’ must have a certain command structure, etc.
[63] UNHCR Study, 76.
[64] Cited in UNHCR Study, 77, fn 317.
[65] Cited in UNHCR Study, 77, fn 318.
[66] UNHCR Study, 78.
[67] UNHCR Statement, 6.
[68] UNHCR Study, 79.
[69] Article 2(c) extends temporary protection to ‘(i) persons who have fled areas of armed conflict or endemic violence; (ii) persons at serious risk of, or who have been the victims of, systematic or generalised violations of their human rights’.
[70] see Council Directive 2001/55/EC of 20 July 2001 on Minimum Standards for Giving Temporary Protection in the Event of a Mass Influx of Displaced Persons and on Measures Promoting a Balance of Efforts between Member States in Receiving Such Persons and Bearing the Consequences thereof [2001] OJ L212/12, art 2(a).
[71] As a comment by the French delegation during the drafting process shows, there is a deep-seated fear that whole populations will flee on the basis of generalized violence if subsidiary protection status does not require individual harm to be demonstrated12199/02 ASILE 45 (25 September 2002) 20 fn 8: the expression ‘international or internal armed conflict’ ‘risks opening the possibility of obtaining subsidiary protection to the entire population of countries involved in conflicts’.
[72] UNHCR Study, 79.
[73] Qualification Directive, art 2(e).
[74] Council of the European Union Presidency Note to Strategic Committee on Immigration, Frontiers and Asylum on 25 September 2002 Doc 12148/02 ASILE 43 (20 September 2002) 5. The Netherlands supported Sweden’s argument that wording from decisions of the Torture Committee should be taken into account to avoid different rulings from different courts of bodies concerning similar situations: 12199/02 ASILE 45 (25 September 2002) 3 fn 3. See also Kacaj* [2001] UKIAT 00018; [2001] INLR 354.
[75] See eg EA v Switzerland (Comm No 28/1995) UN Doc CAT/C/19/D/28/1995 (10 November 1997) para 11.5; X, Y and Z v Sweden (Comm No 61/1996) UN Doc CAT/C/20/D/61/1996 (6 May 1998) para 11.5; IAO v Sweden (Comm No 65/1997) UN Doc CAT/C/20/D/65/1997 (6 May 1998) para 14.5; KN v Switzerland (Comm No 94/1997) UN Doc CAT/C/20/D/94/1997 (19 May 1998) para 10.5; ALN v Switzerland (Comm No 90/1997) UN Doc CAT/C/20/D/90/1997 (19 May 1998) para 8.7; JUA v Switzerland (Comm No 100/1997) UN Doc CAT/C/21/D/100/1997 (10 November 1998) para 6.6; SMR and MMR v Sweden (Comm No 103/1998) UN Doc CAT/C/22/D/103/1998 (5 May 1999) para 9.7; MBB v Sweden (Comm No 104/1998) UN Doc CAT/C/22/D/104/1998 (5 May 1999) para 6.8; KT v Switzerland (Comm No 118/1998) UN Doc CAT/C/23/D/118/1998 (19 November 1999) para 6.5; NM v Switzerland (Comm No 116/1998) UN Doc CAT/C/24/D/116/1998 (9 May 2000) para 6.7; SC v Denmark (Comm No 143/1999) UN Doc CAT/C/24/D/143/1999 (10 May 2000) para 6.6; HAD v Switzerland (Comm No 126/1999) UN Doc CAT/C/24/D/126/1999 (10 May 2000) para 4.10; US v Finland (Comm No 197/2002) UN Doc CAT/C/30/D/197/2002 (1 May 2003) para 7.8.
[76] EA v Switzerland (Comm No 28/1995) UN Doc CAT/C/19/D/28/1995 (10 November 1997) para 11.3.
[77] Report of the Committee against Torture UN GAOR 53rd Session Supp No 44 UN Doc A/53/44 (1998) Annex IX.
[78] v Switzerland (Comm No 28/1995) UN Doc CAT/C/19/D/28/1995 (10 November 1997) para 11.3.
[79] For example, in the UK, the test for a well-founded fear of persecution is a ‘reasonable likelihood’ of such danger: R v Secretary of State for the Home Dept, ex parte Sivakumaran [1987] UKHL 1; [1988] AC 958 (HL). Article 7(b) of the original proposal for the Qualification Directive stated that well-founded fear was to be ‘objectively established’ by considering whether there was ‘a reasonable possibility that the applicant [would] be persecuted’. The Explanatory Memorandum noted that a ‘fear of being persecuted ... may be well-founded even if there is not a clear probability that the individual will be persecuted or suffer such harm but the mere chance or remote possibility of it is an insufficient basis for the recognition of the need for international protection’: See Commission of the European Communities Proposal for a Council Directive on Minimum Standards for the Qualification and Status of Third Country Nationals and Stateless Persons as Refugees or as Persons Who Otherwise Need International Protection COM(2001) 510 final (12 September 2001) 15.
[80] Kacaj* [2001] UKIAT 00018; [2001] INLR 354, para 10.
[81] Kacaj* [2001] UKIAT 00018; [2001] INLR 354, para 15.
[82] See my earlier discussion of this point: J McAdam, Complementary Protection in International Refugee Law (OUP, Oxford, 2007) 61–64.
[83] AIT 2008, para 143.
[84] H Battjes, European Asylum Law and International Law (Martinus Nijhoff Publishers, Leiden 2006) 225, referring also to J-Y Carlier, ‘Réfugiés: Identification et statut des personnes à protéger. La direction “qualification”’, in F Julien-Laferrière, H Labayle and Ö Edström (eds), The European Immigration and Asylum Policy: Critical Assessment Five Years after the Amsterdam Treaty (Bruylant, Brussels, 2005) text to fn 34; R Piotrowicz and C van Eck, ‘Subsidiary Protection and Primary Rights’ (2004) 53 ICLQ 107, 113. Battjes also suggests that the level of risk might be higher than CAT (225 – Carlier 2005, 1B(2)), and also suggests that the HRC jurisprudence is stricter than under the ECHR.
[85] 12199/02 ASILE 45 (25 September 2002) 3 fn 3. See further J McAdam, Complementary Protection in International Refugee Law (OUP, Oxford, 2007) 62–63.
[86] UNHCR Statement, 20.
[87] Li v Canada (Minister of Citizenship and Immigration) [2003] FCJ No 1934; 2003 FC 1514; aff’d in Li v Canada (Minister of Citizenship and Immigration) [2005] FCJ No 1; 2005 FCA 1, paras 18–28.
[88] This test derives from Adjei v Minister of Employment and Immigration [1989] 2 FC 680.
[89] Li v Canada (Minister of Citizenship and Immigration) [2005] FCJ No 1; 2005 FCA 1, paras 18–28. Since this was the interpretation which had been given in Suresh v Canada (Minister of Citizenship and Immigration) [2000] FCJ No 5 (FCA), Justice Rothstein said that Parliament could have enacted a lower test had it desired to depart from that interpretation.
[90] Li v Canada (Minister of Citizenship and Immigration) [2005] FCJ No 1; 2005 FCA 1, para 38.
[91] J Reekie and C Layden-Stevenson, ‘Complementary Refugee Protection in Canada: The History and Application of Section 97 of the Immigration and Refugee Protection Act (IRPA)’ (Part III of the International Association of Refugee Law Judges’ Convention Refugee Status and Subsidiary Protection Working Party First Report, IARLJ 7th World Conference, Mexico City, 2006) 49.
[92] Kacaj* [2001] UKIAT 00018; [2001] INLR 354, para 12.
[93] UNHCR Study, 73.
[94] See Ammari v Sweden App No 60959/00 (22 October 2002); see also references in AIT 2008.
[95] EA v Switzerland UN Doc CAT/C/19/D/28/1995 (10 November 1997): risk must be foreseeable, real and personal. See UNHCR Study, 80.
[96] Kacaj* [2001] UKIAT 00018; [2001] INLR 354, para 12.
[97] UNHCR Study, 79, paraphrasing German FedOff MOI.