Refugee Status Appeals Authority
REFUGEE APPEAL NO. 76044
AT AUCKLAND
Before: RPG Haines QC (Chairperson)
B Dingle (Member)Representing the Appellant: P Thoman
Date of Hearing: 15 and 16 October 2007
Date of Decision: 11 September 2008
DECISION OF THE AUTHORITY DELIVERED BY RPG HAINES QC
INDEX
INTRODUCTION
BACKGROUND
Personal and domestic circumstances
The first claim to refugee status
Recognition of the sons as refugees
The second claim to refugee status - introduction
The second claim to refugee status - particulars
Jurisdiction - second and subsequent claims to refugee status
THE ISSUES
Credibility
COUNTRY INFORMATION “HONOUR” KILLINGS IN TURKEY
Country information - state protection in Turkey
THE RISK ISSUE “A WELL-FOUNDED FEAR”
The legal test
Assessment of the facts
THE HARM ISSUE “BEING PERSECUTED”
The meaning of “being persecuted”
“Being persecuted” and State protection
Protection and the “public - private” divide
Assessment of the facts
THE REASONS FOR THE RISK OF HARM THE CONVENTION GROUNDS
“For reasons of” - the standard of causation
POLITICAL OPINION
Gender and political opinion - introduction
Gender, power and honour crimes
Conclusions on honour crimes
Political opinion - a gender-sensitive interpretation
Political opinion - whether established on the facts
PARTICULAR SOCIAL GROUP
Particular social group - whether established on the facts
THE INTERNAL PROTECTION ALTERNATIVE
Introduction
THE TEXTUAL FOUNDATION FOR THE INTERNAL PROTECTION ALTERNATIVE
The analysis based on protection
The analysis based on risk
Critique of the analysis based on risk
Protection as diplomatic protection
Critique of the “diplomatic protection” interpretation
Conclusion as to textual foundation for the IPA
FORMULATING THE TEST FOR WITHHOLDING RECOGNITION OF REFUGEE STATUS
The “reasonableness” test
The reasonableness test in New Zealand after Butler
Criticisms of the reasonableness test
The protection inquiry - identifying basic norms of civil, political and socio-economic rights
The challenge to the human rights inquiry in Januzi
Januzi and the re-instatement of human rights
Subjectivity revisited - “standards prevailing generally in the country of origin” and “a relatively normal life”
Januzi and AH (Sudan) - conclusions
APPLICATION OF THE INTERNAL PROTECTION ALTERNATIVE TO THE FACTS
CONCLUSION
[1] The appellant is a forty-year-old citizen of Turkey who arrived in New Zealand on 11 February 2000 with her two sons then aged thirteen and fourteen respectively. Her first refugee application filed on 23 February 2000 failed. She now advances a second claim based on her fear of being murdered in a so-called “honour killing”.
[2] In a decision given on 13 April 2007, a refugee status officer declined the new refugee application on credibility grounds. As will be seen, on appeal the Authority has been presented with more cogent evidence than that made available to the refugee status officer and on that evidence has determined that the appellant is to be recognised as a refugee.
Personal and domestic circumstances
[3] The appellant is an Alevi Kurd from the southeast of Turkey. Her family is large, comprising her father, mother, three elder brothers and four sisters. One brother currently lives in Istanbul, the other two in Switzerland where they have been recognised as refugees. A fourth brother who went to Germany took his life approximately five years ago.
[4] Prior to his retirement the appellant’s father was a farmer by occupation and the family was organised on the basis that the father held all the power in the family and the women (particularly the daughters) were relegated to looking after the household and to working on the farm. The appellant received only five years education, the view of her father being that her time was better spent working at home. The father’s authority within the family was absolute and it was common for the appellant, her sisters and mother to be beaten if in his view they were carrying out their tasks inadequately or tardily. The daughters were forbidden from speaking to any male person outside the family.
[5] At seventeen years of age the appellant was told by her father that she was to marry a man from a nearby village whose family were also farmers. Over her protests she was taken one night to that man’s family home and left there. When she begged her newly acquired mother-in-law to send for her parents to take her home, her father arrived to tell her that if she returned to his household he would kill her.
[6] From that point on the appellant lived an unhappy life with a man for whom she had little affection and in whose household she was subject to the control of others, particularly her husband, his father and his mother. She was regularly assaulted by both her husband and by her mother-in-law.
[7] In June 1994 the appellant’s husband went to the United Kingdom where he made an unsuccessful refugee claim. The appellant was given little, if any, information as to the reasons for the husband leaving Turkey. As it turned out, after being detained for some eighteen months, he was deported to Turkey in July 1997. During his absence the appellant and her two sons continued to live in the household of the father-in-law. Three months after the husband returned to Turkey he left once more, arriving in New Zealand in December 1997. At that time the appellant and her sons left the home of her parents-in-law and went to live with her own parents who by this time had moved some distance away to the far south of Turkey.
[8] In March 1999, while geographically separated, the appellant and her husband divorced. However, it had not been the intention of the appellant and her husband that the marriage itself be dissolved. It was a divorce in name only, it being claimed to be a device to shield the appellant from inquiries allegedly being made by the authorities in Turkey about the husband. The precise circumstances in which the husband initiated this “divorce” remain unclear. Be that as it may, on 17 November 1999 the appellant was able to obtain a false Turkish passport on which her two sons were included. On 25 January 2000 all three were issued with a New Zealand visitor’s visa and this was followed by their arrival in New Zealand on 11 February 2000. Thereafter the appellant and her husband resumed living together.
The first claim to refugee status
[9] On 23 February 2000 the appellant’s first refugee claim was filed. The grounds on which that claim were advanced are fully set out in Refugee Appeal No. 73965 (10 February 2005), being the Authority’s decision on that claim. Briefly, the appellant alleged that because of her husband’s active involvement in the People’s Democracy Party (HADEP) he had been forced to flee Turkey. Thereafter the appellant herself had been accused of supporting HADEP and the Kurdistan Workers’ Party (PKK). Even though she had left the household of her parents-in-law to resume living with her own parents, she had been detained and questioned by the authorities on four separate occasions and questioned about her husband’s activities in HADEP. Believing that she was not safe she travelled to New Zealand with her two sons to join her husband.
[10] By this time the husband’s own refugee claim had been partially processed. Following a decline at first instance he had appealed to this Authority. The appeal was heard on 13 March 2000, 30 March 2000, 13 April 2000, 28 April 2000 and 30 October 2000. The appellant was called as a witness at the hearing. In a decision given on 9 November 2000 the then panel of the Authority found both the husband and the appellant to be untrustworthy witnesses. See Refugee Appeal No. 71750 (9 November 2000):
[35] The Authority found the [husband] to be articulate and overbearing, a person whose demeanour suggested strongly that he would stop at nothing and say anything to achieve his goal of refugee status. He prevaricated and attempted to avoid any questions which probed inconsistencies or areas where implausibility was exposed. He resorted to launching into tirades, and presented generally as a devious and untrustworthy witness.[11] On 22 February 2002 the appellant was interviewed at first instance in relation to her own (first) refugee application. In a decision given on 13 June 2002 that application was declined. An appeal to this Authority was lodged, the appeal being heard on 21 January 2005 by a differently constituted panel to that which heard the husband’s appeal. In a decision given on 10 February 2005 the appellant’s appeal was dismissed (Refugee Appeal No. 73965 (10 February 2005)) on credibility grounds:
[36] The [husband’s] wife presented as being cast much in the same mould, and her evidence contained varying inconsistencies and generally, she displayed deviousness comparable with her husband’s.
[34] For the reasons explained in Refugee Appeal No. 71750 it is beyond dispute that all claims made about the husband’s activities with HADEP are not credible. Additionally having seen and heard from the appellant I am left in no doubt that her latest evidence on the same point is totally unreliable. Her answers were vague and implausible and her demeanour highly emotional. The constant refrain when asked for any detail beyond her mere reciting of the claim that several HADEP men used to come and see her husband was “I don’t know”.
[12] In the following month representations were made by the family to the Associate Minister of Immigration seeking his intervention by way of the grant of residence permits but in a decision dated 18 July 2005 he declined the application. On 8 August 2005 the husband and the appellant filed proceedings in the High Court at Auckland challenging the Authority’s decision of 10 February 2005. However, after instructing new counsel, those proceedings were discontinued in June 2006. While those proceedings were in train an unsuccessful approach was made to the Prime Minister.
Recognition of the sons as refugees
[13] On 22 May 2006 the two sons were recognised as refugees by the Refugee Status Branch of the Department of Labour on the grounds that they would be conscripted into the Turkish Army, harshly treated because they are Kurds and required to take part in operations against fellow Kurds, thereby putting them at risk of violating international humanitarian law. In the event of their evading such conscription they would be imprisoned and exposed to the risk of torture.
The second claim to refugee status - introduction
[14] On 22 November 2006 both the husband and the appellant lodged second applications for refugee status. The Authority does not have the husband’s application but in summary the appellant’s application was advanced the following grounds:
(a) Because she left Turkey illegally on a false passport, on return she was at risk of being stopped and questioned by the Turkish authorities.
(b) Both her and her husband’s past history would then be discovered.
(c) Because her two sons are of an age when they should be doing their military service and would not be returning to Turkey with the appellant, the authorities would note their absence and ascribe to the appellant anti-government views.
(d) Her prolonged absence from Turkey would increase the suspicion of the authorities.
[15] For present purposes it is important to note that on the date the appellant and her husband submitted their second refugee claims, they were still living together. That situation was, however, about to change. One week before Christmas of 2006 the appellant’s husband attempted to kill the appellant preparatory to taking his own life. Surviving the attack the appellant terminated what was left of the relationship and she has not seen her husband since. It has been confirmed, however, that he left New Zealand on 12 February 2007 and has not returned.
[16] As can be seen from the earlier narrative, the second refugee claim was a joint enterprise by the appellant and her husband and had every appearance of a meritless, if not abusive, claim.
[17] However, by the time of the first instance interview on 13 February 2007, the attempt on the appellant’s life had taken place and there was now added to her second refugee claim her fear that should she return to Turkey, she would be killed by her husband or his family or by the appellant’s own family.
[18] As will be seen, it is on this latter ground alone that the claim to refugee status succeeds. The other grounds set out in the second claim do not, either individually or taken together, establish a credible basis for a claim to refugee status and for that reason will not be examined or discussed.
The second claim to refugee status - particulars
[19] The evidence in support of the claim that the appellant is at risk of being killed by family members should she return to Turkey needs to be expanded on.
[20] On the appellant joining her husband in New Zealand in February 2000 his strict control over her life and activities continued. Although she was granted a work permit pending the determination of her first refugee claim, she was forbidden by her husband both from working and from attending English language classes. He told her that should she be allowed out of the home he believed she would socialise with English-speaking people and find another man. Even the clothes she wore were dictated by the husband. If, for example, she wore a dress with short sleeves she would be required to change or to wear something on top, even during the warm and humid summer months in Auckland. On a regular basis she was subjected to violence, abusive language and psychological humiliation.
[21] In 2006, following the decline of their respective refugee claims and their representations to the Minister of Immigration (and the futile judicial review proceedings), the husband’s behaviour towards the appellant became even more extreme as his psychological condition deteriorated. By then the family was destitute, the work permits having been withdrawn consequent on the decline of the first refugee claims. From April 2006, no longer able to afford to pay rent, the appellant and her husband moved into a garage at the home of friends who themselves are Kurdish refugees from Turkey. The appellant’s sons slept in the family vehicle parked outside. The wife of the family in whose garage the appellant and her husband were living, Ms [H], was called as a witness and described how the husband’s mental condition declined during the year as he grew increasingly depressed day by day. She was also witness to the appellant being insulted, sworn at and assaulted by her husband. The appellant confirmed the deterioration in her husband’s mental state and told the Authority that he would not take medication which had been prescribed for him. He repeatedly accused her of having found a “boyfriend” and threatened to kill her (the appellant) preparatory to his own suicide.
[22] One night, a week before Christmas of 2006, while asleep in the garage, the appellant woke to find her husband trying to strangle her. When she asked what he was doing he said that he was going to kill her and then take his own life. He had already prepared a rope which she could see hanging from a beam in the garage. In the struggle which ensued the appellant was able to escape the garage and to reach the house where she took refuge with Ms [H]. During the confrontation which followed the appellant removed her wedding ring and threw it at her husband, telling him that the marriage was over and that he was to leave. That was the last time the appellant saw her husband and as mentioned he left New Zealand on 12 February 2007 and returned to Turkey.
[23] In her evidence to the Authority, Ms [H] said that on the day following this incident the appellant’s husband telephoned Ms [H]’s home (the appellant herself did not have a telephone in the garage or a cellphone) accusing his wife of having a new boyfriend and stating that it was his intention to kill her (the appellant).
[24] Also on the day following the incident the husband’s family in Turkey telephoned Ms [H]’s home four times and spoke to the appellant. They told her that they had learnt from the husband that she had terminated the marriage. They said that such a thing had never happened before in their family and was not permitted by the Kurdish community. They told the appellant that they (the husband’s family) were required to “cleanse their dignity” as they had been shamed in front of the entire Kurdish community. The appellant’s father-in-law, who holds absolute authority within the husband’s family, said that he would personally ensure that the appellant was killed. Similar telephone calls were made to the appellant once a week thereafter, sometimes once every five days. In total some ten to seventeen such telephone threats were made and the appellant became frightened. She also received a telephone call from her husband before he returned to Turkey in which he advised that should she return to Turkey he would cleanse his shame by killing her.
[25] Both of the appellant’s sons gave evidence at the appeal hearing. Each has witnessed the abuse to which their mother has been subjected and they themselves have been victims. One son in particular described how his father had broken his (the son’s) finger and the other recalled receiving “four hidings”. Both had been strictly controlled and disciplined by their father. The youngest son also described how he had remained in telephone contact with his paternal grandparents on a monthly basis. Since December 2006 he had been repeatedly instructed by them to tell his mother to return to Turkey so that his father could do his duty to “clean up the mess” following his betrayal by the appellant. It was a matter of honour that he do so.
[26] The evidence of the appellant was that since June 2007 her own family have threatened her life after the husband contacted them in Turkey and told them of the separation. The appellant had until then concealed the ending of the relationship. However, once her family heard that she had left her husband, her father and the brother who lives in Istanbul had threatened to kill her, saying that they were ashamed of her and that they did not want her back in Turkey.
[27] This evidence was confirmed by Ms [H] who told the Authority that some three to four months prior to the hearing before the present panel she had begun to receive telephone calls from persons who introduced themselves as the appellant’s father and brother. They told her (Ms [H]) that they had learnt from the appellant’s husband that the appellant was “a bad woman” and that on her return to Turkey they would kill her. Ms [H] told them that she did not want them to call her and by the month prior to the appeal hearing there had been no further telephone calls.
[28] In summary the appellant’s claim to refugee status is based on evidence that having terminated her relationship with an increasingly unstable, jealous and violent husband she is now at risk in Turkey of being killed not only by her husband and his family, but also by her own family in order to maintain their “honour”.
Jurisdiction - second and subsequent claims to refugee status
[29] In only limited circumstances can a second or subsequent claim to refugee status be made. Those circumstances are prescribed by s 129J(1) of the Immigration Act 1987. The claimant must show that since the determination of the first refugee claim circumstances in his or her home country have changed to such an extent that the further claim is based on significantly different grounds to the previous claim:
129J. Limitation on subsequent claims for refugee status—
(1) A refugee status officer may not consider a claim for refugee status by a person who has already had a claim for refugee status finally determined in New Zealand unless the officer is satisfied that, since that determination, circumstances in the claimant's home country have changed to such an extent that the further claim is based on significantly different grounds to the previous claim.
(2) In any such subsequent claim, the claimant may not challenge any finding of credibility or fact made in relation to a previous claim, and the officer may rely on any such finding.
[30] A person whose subsequent claim has been declined by a refugee status officer, or whose subsequent claim has been refused to be considered by such officer on the grounds that the statutory criteria have not been satisfied, may appeal to this Authority. Section 129O(1) provides:
129O. Appeals to Refugee Status Appeals Authority—
(1) A person whose claim or subsequent claim has been declined by a refugee status officer, or whose subsequent claim has been refused to be considered by an officer on the grounds that circumstances in the claimant's home country have not changed to such an extent that the subsequent claim is based on significantly different grounds to a previous claim, may appeal to the Refugee Status Appeals Authority against the officer's decision.
[31] Jurisdiction to hear a subsequent claim is the corollary of the fact that the Refugee Convention does not require that the individual leave his or her country on account of a well-founded fear of being persecuted. The individual may decide to ask for recognition of his or her refugee status after having already been abroad for some time. A person who was not a refugee when he or she left the country of origin, but who becomes a refugee at a later date, is called a refugee sur place. See the UNHCR Handbook on Procedures and Criteria for Determining Refugee Status at para [94]. Refugee status is declaratory, not constitutive, and the obligation of non-refoulement attaches to anyone who satisfies the Convention definition at the date of the proposed removal from New Zealand.
[32] On the facts, jurisdiction to determine the appellant’s second claim to refugee status has been clearly established. The events which took place in Auckland in December 2006 and the subsequent threats to her life mean that the appellant now asserts a risk of serious harm in Turkey which did not exist prior to December 2006. Clearly circumstances in her home country have changed to such an extent that it can be properly said that the second claim to refugee status is based on significantly different grounds to the previous claim.
[33] Jurisdiction having been established, it is now possible to address the question whether, on the evidence now before the Authority, the appellant has established that she has a well-founded fear of being persecuted for reasons of her race, religion, nationality, membership of a particular social group or political opinion.
[34] The Inclusion Clause in Article 1A(2) of the Refugee Convention relevantly provides that a refugee is a person who:
“... owing to well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country; or who, not having a nationality and being outside the country of his former habitual residence is unable or, owing to such fear, is unwilling to return to it.”
[35] In terms of Refugee Appeal No. 70074/96 Re ELLM (17 September 1996); [1998] NZAR 252, 258-263 the principal issues are:
1. Objectively, on the facts as found, is there a real chance of the appellant being persecuted if returned to the country of nationality?
2. If the answer is Yes, is there a Convention reason for that risk of being persecuted?
[36] If, and only if, both of these issues are answered in the affirmative will the Authority need to address the issue whether the appellant is nevertheless not to be recognised as a refugee by reason of the fact that she can access meaningful state protection should she be returned to Turkey.
Credibility
[37] Before the identified issues can be addressed an assessment must be made of the appellant’s credibility. Any refugee claimant whose credibility as a witness has been comprehensively rejected by two successive panels of this Authority faces an almost insurmountable hurdle in establishing her credibility on her third appearance before a new panel of the Authority.
[38] The present panel is mindful of the need to avoid the re-litigation of issues thoroughly traversed by the two earlier panels while keeping an open mind in relation to the credibility of any new evidence advanced by the appellant in support of this, her second refugee claim. Applying s 129P(9) of the Act, we have decided to rely on the findings of the two earlier panels in relation to the circumstances advanced in support of the first refugee claim but to assess for ourselves any new evidence relevant to the sur place basis of the claim, the first and second claims being based on entirely different facts. In the result, the appeal stands or falls on the claim that the appellant is at risk of being killed by her husband and family members should she return to Turkey.
[39] When giving evidence about her own life experiences and freed of the need to support the false claim to refugee status by her husband, the appellant was a credible and persuasive witness. Our favourable assessment of her credibility has been assisted by the independent evidence of Ms [H] who gave the Authority no reason to doubt her account of the events which unfolded at her home once the appellant and her husband took up residence there in the garage. Furthermore, the appellant’s two sons (now aged 21 and 20 respectively), both gave frank accounts of their home life and of the violence they and their mother have suffered at the hands of their father.
[40] In making our credibility assessment we have not found it helpful to hold against the appellant the fact that on the night of the attempt on her life she dissuaded Ms [H] from calling the police. It is well-established that there are many reasons why a woman who for years has been the victim of control and domestic violence will not seek the intervention of the state. In addition the appellant has little formal education, has been cloistered for most of her life, does not speak English and has been prevented by her husband from engaging with or integrating into New Zealand society. Her understanding as at December 2006 of her rights and remedies is best described as minimal. She believed that by calling the police she would be making a bad situation worse, a judgment which we are ill-equipped to second-guess.
[41] We have accordingly concluded that the appellant has given a truthful account of the events which have occurred in New Zealand subsequent to her arrival here, corroborated as she is in this respect by her sons and by Ms [H]. We accept the central elements of the claim, namely:
(a) That consequent upon the appellant terminating the marital relationship, the husband believes he is duty bound to cleanse both his “honour” and that of his family by killing the appellant.
(b) That the husband’s family will ensure that either the husband or someone else in the family kills the appellant.
(c) That the appellant’s own father as well as her brother who lives in Istanbul feel equally bound by duty to kill the appellant in order to restore the “honour” of their family.
[42] However, these findings on their own are insufficient to allow an assessment to be made as to whether the appellant has established the four definitional elements critical to this case, namely risk (ie well-foundedness), harm (ie being persecuted), the reason for that risk (ie whether a Convention ground is present) and a failure of state protection. For these elements to be assessed the human rights situation in Turkey is of critical importance. The relevant country information is examined next.
[43] The country information on so-called “honour” killings in Turkey is substantial and it is not intended to provide here a comprehensive survey.
[44] Violence against women, including honour killings and rape continue to be a widespread problem in Turkey: United States Department of State, Country Reports on Human Rights Practices for 2007: Turkey (March 2008). Despite the efforts of the government, honour-related crimes in Turkey show little sign of abating. It is reported in “A Dishonourable Practice” The Economist (April 14, 2007) 57 that a parliamentary report of August 2006 found that 1,091 such crimes had been committed in the past five years, which is over four a week. Only three of 51 honour killers interviewed for another study said they had any regrets.
[45] Many of the murders occur in the Kurdish provinces where there is a deeply entrenched patriarchal and feudal system. Rampant poverty and illiteracy have been exacerbated by the forced eviction of millions of Kurdish villagers by the army in its war against PKK rebels in the 1990s. But honour crimes are not a uniquely Kurdish phenomenon nor are they properly associated with any particular society or religion. See Dr Dicle Kogacioglu, “The Tradition Effect: Framing Honor Crimes in Turkey” (2004) 15 Differences 118, 130; Dr Radhika Coomaraswamy, “Identity Within: Cultural Relativism, Minority Rights and the Empowerment of Women” (2002) 34 Geo. Wash. Int’l L. Rev 483, 495-496; Human Rights Watch, The Human Rights Watch Global Report on Women’s Human Rights (1995) 355 and Lynn Welchman and Sara Hossain (eds), “Honour”: Crimes, Paradigms and Violence Against Women (2005, Zed Books). Dr Dicle Kogacioglu at op cit 129-130 makes the point in the following terms:
The single clear fact about the suicides is their location: south-eastern Turkey. Since honor killings are thought to be most prevalent in the Southeast, the JDP [Justice and Development Party] collapses the two. In this way, the party contributes to the mainstream view that honor crimes are a phenomenon contained in the Southeast, a view held by people of various political orientations. In the absence of national and regional statistics about honor crimes, it is difficult to assess this assumption, although media reports in other parts of the country make it seem hardly realistic. And even if killings per se are more prevalent in the Southeast, other practices that can be placed under the rubric of honor crimes, such as the limits imposed on women’s rights to travel or to receive education, and other forms of violence women face in the name of the protection of honor, are fairly commonplace even in cosmopolitan centers such as Istanbul, Ankara, and Izmir.
To single out the Southeast implies that honor crimes are primarily a Kurdish phenomenon, as the area is populated primarily by Kurds. Such a portrayal amounts to the stigmatization of Kurds, a process that is already very much under way due to past armed conflicts between the Kurdish guerilla forces and the Turkish military, with the ensuing forced migration and poverty of Kurds. Here we see the ethnicization of the tradition effect: honor crimes attributed to the traditions of an already disadvantaged ethnic group and its region. This enables other parts of the country to be imagined as somehow immune to the problem.
These observations have particular relevance to the Convention grounds shortly to be discussed as well as to the issue of the internal protection alternative. In this context it is helpful to note the further point made by Dr Dicle Kogacioglu at op cit 119 that honour crimes stand at the intersection of multiple political and social dynamics.
[46] The United States Department of State, Country Reports on Human Rights Practices for 2007: Turkey (March 2008) provides the following general summary:
The law prohibits rape, including spousal rape; however, the government did not effectively enforce the law. Victims often waited days or weeks to report incidents for fear of embarrassment or reprisals, which hindered the possibility of effective prosecution of assailants. Experts worked during the year to convince the government to accept psychiatric victim reports as alternative forms of evidence. Cases of rape were underreported.
Violence against women, including spousal abuse, was a serious and widespread problem. The law prohibits violence against women, including spousal abuse. The government did not effectively enforce the law...
...
Women’s NGOs reported that more than 150,000 women were victims of domestic violence between 2001 and 2005. The government continued to show slow progress on implementing a 2004 law stipulating the need for shelters for women victims of domestic violence in towns with a population of more than 50,000. According to the government, its institution for Social Services and Orphanages operated 23 shelters for female victims of domestic violence and rape with a total capacity of 405. The government reported that provincial government offices, municipalities and NGOs operated 18 shelters, and that one private foundation operated a shelter.
KA-MER, the leading women’s organization in the southeast, reported that from 2003-2007 a total of 198 women from eastern and southeastern Anatolia contacted KA-MER to report that their family had threatened them with honor killings. Of these cases, three of the women died from injuries sustained in the attacks, one committed suicide, and 27 were pressured to commit suicide. The father or husband decided the fate of the woman in the vast majority of the cases. The report observed that 76 of these “decision makers” were illiterate, while 47 had no education beyond junior high school. Increased education levels correlated with a drop in the rate of such crimes. “Disobedience” was determined to be the most common reason given to justify honor killings. Disobedience was variously defined as refusing to marry the person the family had chosen, refusing to have sex with a brother-in-law or father, not agreeing to prostitute oneself, not fulfilling the demands of husbands, fathers, brothers, or other elders, and interrupting man-to-man conversations.
The government reported that there were 37 victims of honor killings during the year and 1,806 honor killings between 2001 and 2006. During the same period, 5,375 women committed suicide. After the government increased penalties for honor killings in 2005, family members increasingly pressured girls to kill themselves in order to preserve the family’s honor, according to women’s rights groups. Government officials worked with advocacy groups such as KA-MER to hold town hall meetings and set up rescue teams and hot lines for endangered women and girls. Under the law, honor killings required punishment of life imprisonment. Women’s rights groups reported that there remained dozens of such killings every year, mainly in conservative Kurdish families in the southeast or among migrants from the southeast living in large cities. Because of sentence reductions for juvenile offenders, observers noted that young male relatives often were designated to perform the killing.
The summary goes on to note that it was not until 13 November 2006 that a court in Istanbul ordered the first life sentence for an honour killing.
[47] Much of the country information confirms that honour-related crimes in Turkey are substantially under-reported and it is said that the State Institute of Statistics does not collect statistical data on honour crimes: Dr Dicle Kogacioglu, op cit 125. The official annual figure is 70 but the parliamentary report of August 2006, which says that 1,091 honour crimes have been committed in the past five years, is premised on a death rate of over 200 per annum. Academics also put the number in the hundreds. Were honour suicides also to be taken into account the annual death rate would be even higher. See Helena Smith, “Hundreds Die in Turkey as Honour Suicides Replace Killings”, The Guardian Weekly (31 August 2007) 3 where it is reported that in Batman, a city of 250,000 situated in the southeast of Turkey, more than 300 women have attempted suicide since 2001. Seven died in one month alone. Smith reports that similar death rates are seen all over the southeast. Women’s groups and human rights advocates believe the suicides are in fact honour killings passed off as suicides, the women being forced to take their own lives:
Sometimes adultery, or a wish for divorce, prompts an all-male family council to order a killing. But the list of “offences” is long: rape, incest, pregnancy caused by both, ringing a radio chat show, exchanging eye contact with a boy or wearing skimpy clothing.
[48] So alarming is the phenomenon of honour suicides that in May 2006 Dr Yakin Ertürk, the UN Special Rapporteur on Violence Against Women, its Causes and Consequences and herself a citizen of Turkey, went on “mission” to investigate the suicides of women in eastern and southeastern Turkey and the claims that the deaths of these women may be instances of murder or forced suicide. Her conclusions are to be found in the Report of the Special Rapporteur on Violence Against Women, Its Causes and Consequences: Mission to Turkey (22-31 May 2006) (A/HRC/4/34/Add.2 (5 January 2007)) (the Ertürk report). The summary to the report records that the suicides occurring in the southeastern/eastern region of Turkey are “intimately linked to violence emanating from the understandings of honour and customary law:
There are reasonable grounds to assume that some of the recorded suicide cases are indeed disguised murders. In other cases, family members appear to have instigated the suicide.
[49] Addressing the underlying reasons for the suicides Dr Ertürk concludes at para [47] of the body of the report that:
... the rigid patriarchal oppression of women and the human rights violations that go along with it - especially forced and early marriages, domestic violence, incestuous rape and denial of reproductive rights - [are] key contributing factors to the suicides of women in the region. In the absence of adequate state protection, suicide may be the only option for women to escape extreme violence and oppression.
Country information - state protection in Turkey
[50] On the question of state protection, it is convenient to begin with the June 2004 report by Amnesty International, Turkey: Women Confronting Family Violence (AI Index: EUR44/013/2004, 2 June 2004) which at 18 and 19 records that violence against women is widely tolerated and even endorsed by community leaders and at the highest levels of the government and judiciary. The authorities rarely carried out thorough investigations into women’s complaints about violent attacks or murders or apparent suicides of women:
At every level of the criminal justice system, the authorities failed to respond promptly or rigorously to women’s complaints of rape, sexual assault or other violence within the family. The police are reluctant to prevent and investigate family violence, including the violent deaths of women. Their own record of human rights violations makes victims of domestic violence afraid to seek their help. Prosecutors refuse to open investigations into cases involving domestic violence or to order protective measures for women at risk from their family or community. The police and the courts do not ensure that men who are served with court orders, including protection orders, comply with them.
[51] The Amnesty report goes on to record that men are accorded undue leniency in sentencing on grounds of “provocation” by their victim and on the flimsiest of evidence. At p 19 the barriers facing women in accessing protection from violence are addressed:
There are many barriers facing women who need access to justice and protection from violence. Police officers often believe that their duty is to encourage women to return home and ‘make peace’ and fail to investigate the women’s complaints. Many women, particularly in rural areas, are unable to make formal complaints, because leaving their neighbourhoods subjects them to intense scrutiny, criticism and, in some cases, violence. Women in Kurdish and Arabic speaking areas of the country may not be able to communicate well in Turkish, and may fear further violence at the hands of the police or security forces.
[52] However, this report was published four years ago and account must be taken of such changes as may have occurred since then consequent on Turkey’s ambition to become a member of the European Union. Following the decision of the Helsinki European Council of December 1999 to grant the status of candidate country to Turkey, accession negotiations with Turkey were opened in October 2005 and legal reforms followed, with the Commission of the European Communities reporting regularly to the Council and the European Parliament on the progress being made by Turkey in preparing for EU membership. See for example the 2006 and 2007 progress reports published by the Commission, namely Turkey 2006 Progress Report (Brussels, 08.11.2006 SEC (2006) 1390) and Turkey 2007 Progress Report (Brussels, 6.11.2007 SEC (2007) 1436). For an academic commentary on the reforms, reference can be made to Rebecca E Boon, “They Killed Her for Going Out With Boys: Honor Killings in Turkey in Light of Turkey’s Accession to the European Union and Lessons for Iraq” (2006) 35 Hofstra L. Rev. 815.
[53] The Commission’s Turkey 2006 Progress Report notes at p 17 that notwithstanding some legislative improvements, crimes in the name of honour and suicides committed by women due to the influence of the family continue to occur. While describing the legal framework as “overall satisfactory”, the Report records that “implementation remains a challenge”. The Turkey 2007 Progress Report at p 18 also records that domestic violence against women continues to be widespread and that honour killings, early and forced marriages continue to occur. In terms which echo the 2006 report, the conclusion reached in the 2007 report is that:
Overall, progress has been achieved on protecting women from violence. The legal framework guaranteeing gender equality is in place. However, further efforts are needed to translate it into social reality.
[54] The substantial gap between the legal reforms and “social reality” is also one of the points made in the assessment made by Dr Yakin Ertürk in her report of 5 January 2007 at para [59]:
While from a gender perspective the penal law framework is now overall adequate, significant problems persist in its implementation. Many politicians and administrators still regard domestic violence and forced marriages as internal family matters in which the State should not intervene. In the few cases reported to the law enforcement authorities, officials often try to broker an agreement between parties of inherently unequal power instead of protecting the victims and prosecuting their perpetrators. Particularly in the context of the volatile political situation in the region, some officials are also prone to tacitly subordinate State law to customary forms of conflict resolution in order not to upset relations with local power structures.
The Summary to the Ertürk report contains the following statement:
On 1 June 2005, groundbreaking reforms to the Turkish Penal Code entered into force, successfully removing the most obvious patriarchal biases from the law. Despite these advances in the legislative framework, many problems persist in their actual implementation, including lack of sufficient protective mechanisms for women such as shelters.
[55] It would therefore appear that while changes to the legal landscape in Turkey have taken place since mid-2005, the 2004 Amnesty International assessment continues to be largely valid today. The point is underlined by the United States Department of State, Country Reports on Human Rights Practices for 2007: Turkey already referred to which records that while the law prohibits violence against women, including spousal abuse, the government does not effectively enforce the law. In addition, the government continued to show slow progress in implementing a 2004 law stipulating the need for shelters for women victims of domestic violence in towns with a population of more than 50,000.
[56] Against this background it is now possible to address the issues of risk, harm, Convention grounds and state protection.
The legal test
[57] In the Authority’s jurisprudence the well-founded standard has been understood as mandating the establishment of a real chance of being persecuted. See for example Refugee Appeal No. 72668/01 [2002] NZAR 649 at paras [111] to [154]. The standard is an entirely objective one. The trepidation of the refugee claimant, no matter how genuine or intense, does not alter or affect the legal standard and is irrelevant to the well-foundedness issue. Any subjective fear of harm, while relevant to the question whether the claimant is unable or unwilling to avail him or herself of the protection of the country of nationality, is of no relevance to whether the anticipation of being persecuted is well-founded. See Refugee Appeal No. 75692 [2007] NZAR 307 at paras [76] to [90] and the Michigan Guidelines on Well-Founded Fear (2005) 26 Mich. J. Int’l L. 491.
Assessment of the facts
[58] We are satisfied that there is a real chance of the appellant being killed or otherwise seriously harmed by her husband or by family members. Both families have for generations been organised on strict patriarchal lines. The appellant’s personal experiences in relation to her education, marriage and assigned gender role in the households of her father and of her husband evidence an unwavering and rigid adherence to so-called “tradition”, as that term is currently understood in rural families from southeastern Turkey. The country information establishes that violence against women is pervasive, honour killings and “honour suicides” common and that there is a lack of protection mechanisms. There is little doubt that should the appellant return to the household of her parents (presently in the far south of Turkey) or that of her husband and parents-in-law (presently in the southeast) she will either be killed or suffer serious harm. The same applies should she take up residence in Istanbul where one of her brothers lives, he having expressed his intention to kill her.
[59] In the result, the well-founded element of the Convention is satisfied. This finding will have some significance in the context of the later discussion of whether there is an internal protection alternative in Turkey for the appellant.
[60] Next it is necessary to determine whether the risk of harm faced by the appellant is a risk of “being persecuted”.
The meaning of “being persecuted”
[61] The Authority has for many years interpreted the “being persecuted” element of the refugee definition as the sustained or systemic violation of basic human rights demonstrative of a failure of state protection. In other words, core norms of international human rights law are relied on to define the forms of serious harm which are within the scope of “being persecuted”. This is sometimes known as the human rights understanding of “being persecuted” and is fully explained in Refugee Appeal No. 74665/03 [2005] NZAR 60; [2005] INLR 68 at paras [36] to [125]. For specific case law on gender-related persecution reference may be made to Refugee Appeal No. 2039/93 Re MN (12 February 1996) at pp 14-16, 19-40 and Refugee Appeal No. 71427/99 [2000] NZAR 545; [2000] INLR 608 at paras [43] to [82] and [111] to [120]. In the first of these cases the risk of being killed by an honour crime was accepted as falling within the Refugee Convention. That decision has been consistently applied since 1996 at first instance in other honour crime cases.
“Being persecuted” and State protection
[62] As noted in Refugee Appeal No. 74665/03 at para [51], central to the definition of the term “refugee” is the concept of state protection. Consequently the phrase “being persecuted” must be interpreted within the wider framework of the failure of state protection. Both in New Zealand and in the United Kingdom it is settled that the determination whether the particular facts establish a well-founded risk of being persecuted requires identification of the serious harm faced in the country of origin and an assessment of the state’s ability and willingness to respond effectively to that risk. “Being persecuted” is therefore to be seen as the construct of two separate but essential elements, namely the risk of serious harm and a failure of state protection. This has been expressed in the formula Persecution = Serious Harm + The Failure of State Protection: R v Immigration Appeal Tribunal; Ex parte Shah [1999] 2 AC 629 (HL) at 653F.
[63] As to the standard of state protection, the Authority in Refugee Appeal No. 71427/99 at para [66] held that the issue is whether the protection available from the state will reduce the risk of serious harm to below the level of well-foundedness, or, as it is understood in New Zealand, to below the level of a real chance of serious harm. This is a more exacting standard than that which is applied in the United Kingdom subsequent to Horvath v Secretary of State for the Home Department [2001] 1 AC 489 (HL) but has the distinct advantage of securing compliance with the non-refoulement obligation in Article 33 of the Convention: Refugee Appeal No. 74665/03 (7 July 2004); [2005] NZAR 60; [2005] INLR 68 at [54].
Protection and the “public - private” divide
[64] The refugee definition does not require that the state itself be the agent of harm. Persecution at the hands of “private” or non-state agents of persecution equally falls within the definition: Refugee Appeal No. 2039/93 Re MN (12 February 1996) at p 17 and Refugee Appeal No. 71427/99 [2000] NZAR 545; [2000] INLR 608 at paras [56] to [61]. As noted at para [60] of that decision, there are four situations in which it can be said that there is a failure of state protection:
(a) Persecution committed by the state concerned.
(b) Persecution condoned by the state concerned.
(c) Persecution tolerated by the state concerned.
(d) Persecution not condoned or not tolerated by the state concerned, but nevertheless present because the state either refuses or is unable to offer adequate protection.
The appellant’s case is that the failure of state protection flows from a combination of (b), (c) and (d) above.
Assessment of the facts
[65] In the present case the rights in question are clear and straightforward. There is no need for an extensive analysis. The appellant’s case is based on an anticipation of being killed or seriously harmed by non-state agents, being her husband, his family, her own father and her brother who lives in Istanbul. Should either form of harm eventuate, it would be a violation of her right to life and her right not to be subjected to torture or to cruel, inhuman or degrading treatment or punishment (Articles 6 and 7, International Covenant on Civil and Political Rights, 1966). Even assuming that the infliction of such harm will be neither condoned nor tolerated by the State in Turkey, the risk of serious harm will nevertheless be present because, as the country information shows, the State either refuses or is unable to offer adequate protection to women exposed to the kind of predicament faced by the appellant. It follows that the “being persecuted” element of the refugee definition is established.
[66] It is now possible to turn to the Convention grounds.
[67] A well-founded fear of being persecuted on its own is insufficient to establish refugee status. Protection under the Refugee Convention requires the risk of harm to be “for reasons of” one of the five Convention grounds, namely race, religion, nationality, membership of a particular social group or political opinion.
“For reasons of” - the standard of causation
[68] As pointed out by the Authority in Refugee Appeal No. 72635/01 (6 September 2002); [2003] INLR 629 at paras [162] to [179], the “for reasons of” clause of the refugee definition relates not to the word “persecuted” but to the phrase “being persecuted”. The employment of the passive voice (“being persecuted”) establishes that the causal connection required is between a Convention ground and the predicament of the refugee claimant. The Convention defines refugee status not on the basis of a risk “of persecution” but rather “of being persecuted”. The language draws attention to the fact of exposure to harm, rather than to the act of inflicting harm. The focus is on the reasons for the claimant’s predicament rather than on the mind set of the agent of persecution. In this context the Authority has held that it is sufficient for the refugee claimant to establish that the Convention ground is a contributing cause to the risk of “being persecuted”. It is not necessary for that cause to be the sole cause, main cause, direct cause, indirect cause or “but for” cause. It is enough that a Convention ground can be identified as being relevant to the cause of the risk of being persecuted. In this regard the Authority’s jurisprudence and the Michigan Guidelines on Nexus to the Convention Ground (2002) 23 Mich. J. Int’l L. 210 are in accord.
[69] On the facts, two Convention grounds are relevant to the appellant’s case, namely political opinion and membership of a particular social group. As will be seen, they are not in the present context mutually exclusive grounds and comfortably overlap. The social group ground does have the advantage of enjoying support from recent developments in refugee jurisprudence but we believe that the political opinion ground is too often overlooked in cases involving gender-related persecution. Because our finding is that the political opinion ground is, on the facts, the preferable of the two available grounds, we intend addressing that ground first. We begin with an examination of gender in the context of claims to refugee status based on the political opinion ground.
Gender and political opinion - introduction
[70] Neither the refugee definition nor the Convention in general refers to sex or gender.1 This omission is, however, without consequence. The ordinary meaning of Article 1A(2) in its context and in the light of the object and purpose of the Convention requires the conclusion that the Convention protects both women and men and that it must therefore be given a gender-inclusive and gender-sensitive interpretation.
[71] Equally integral are the power structures in the country of origin and in particular the civil, political, social, and economic position of the refugee claimant. Gender is a primary way of signifying relationships of power: Heaven Crawley, Refugees and Gender: Law and Process (Jordans, 2001) 6-7:
The term “gender” ... refers to the social construction of power relations between women and men, and the implications of these relations for women’s (and men’s) identity, status, roles and responsibilities (in other words, the social organisation of sexual difference). Gender is not static or innate but acquires socially and culturally constructed meaning because it is a primary way of signifying relations of power. Gender relations and gender differences are therefore historically, geographically and culturally specific, so that what it is to be a “woman” or “man” varies through space and over time. Any analysis of the way in which gender (as opposed to biological sex) shape the experiences of asylum-seeking women must therefore contextualise those experiences.
[72] The relevance of power structures to gender-based claims is often overlooked. Unthinking application of the seminal decision in R v Immigration Appeal Tribunal; Ex parte Shah [1999] 2 AC 629 (HL) [that gender may be the defining characteristic of a particular social group] can lead to the unintended result that all cases involving gender-related persecution are conceived as “particular social group” cases and that if the female refugee claimant cannot bring herself within this Convention ground, the refugee claim must fail. The risk is real. A recent survey of forty-two countries covered by the Europe Bureau of the UNHCR reveals that nearly three-quarters of European countries have not recognised the gendered nature of political activities in countries of origin: Heaven Crawley and Trine Lester, Comparative Analysis of Gender-Related Persecution in National Asylum Legislation and Practice in Europe (UNHCR Evaluation and Policy Analysis Unit, Department of International Protection, and Regional Bureau for Europe, EPAU/2004/05, May 2004) at para [300]. Addressing the danger of the social group ground obscuring that of political opinion, the study recommended at para [443] that:
States should refrain from assuming that it is appropriate to categorise all gender-related persecution claims under the PSG [particular social group] ground. To do so may limit the development of a gender-sensitive interpretation of all Convention grounds, and effectively marginalise gender-related claims by ignoring the social and political context in which women’s experiences take place.
This recommendation has been formally incorporated into the subsequent UNHCR Handbook for the Protection of Women and Girls (January 2008) at para [4.2.6]:
... there may be a tendency to examine claims solely under the “particular social group” ground of the refugee definition. Women and girls who oppose harmful practices and violence which violate their rights can, for instance, also be seen as facing persecution on account of their political opinion.
[73] This point has been acknowledged in the Authority’s jurisprudence which recognises that because it is possible for Convention grounds to overlap, it is best to identify the principal or strongest ground (or grounds) in relation to which the “for reasons of” inquiry is to be conducted: Refugee Appeal No. 71427/99 at para [90]. Expressed more succinctly by Maryellen Fullerton in “A Comparative Look at Refugee Status Based on Persecution Due to Membership in a Particular Social Group” 26 Cornell Int’l L. J. 505, 550 (1993) (cited with approval in Refugee Appeal No. 2039/93 Re MN (12 February 1996) at p 50), it is just as important to prevent a claim based on political opinion to masquerade as persecution based on social group, as it is to prevent a social group claim to masquerade as a political opinion case. See also Thomas Spijkerboer, Gender and Refugee Status (Ashgate, 2000) at 169-170.
Gender, power and honour crimes
[74] Identification of the most appropriate Convention ground or grounds requires the decision-maker to go beyond the information relevant to risk. So, for example, in the present case which is about the prevalence and location of honour killings in Turkey, an analysis is required of what it means to talk of honour killings. “Culture” and “tradition” are not apolitical, nor are they detached from the prevailing power relations and the economic and social circumstances in which they operate: Ertürk report at para [62]. Every deployment of tradition has its effects in terms of the distribution of power; honour crimes stand at the intersection of multiple political and social dynamics: Dr Dicle Kogacioglu op cit at 119-120.
[75] The general point is explained by Dr Radhika Coomaraswamy, the 1994-2003 UN Special Rapporteur on Violence against Women, its Causes and Consequences in her Preface entitled “Violence against Women and ‘Crimes of Honour’” in Lynn Welchman and Sara Hossain (eds), “Honour”: Crimes, Paradigms and Violence Against Women (2005, Zed Books) at xi:
... violence against women is closely linked to the regulation of sexuality. ... In many societies the ideal of masculinity is underpinned by a notion of ‘honour’ - of an individual man, or a family or a community - and is fundamentally connected to policing female behaviour and sexuality. Honour is generally seen as residing in the bodies of women. Frameworks of ‘honour’, and its corollary ‘shame’ operate to control, direct and regulate women’s sexuality and freedom of movement by male members of the family. Women who fall in love, engage in extramarital relationships, seek a divorce, or choose their own husbands are seen to transgress the boundaries of ‘appropriate’ (that is, socially sanctioned) sexual behaviour. ‘Regulation’ of such behaviour may in extreme cases involve horrific direct violence - including ‘honour killing’, perhaps the most overt example of the brutal control of female sexuality - as well as indirect subtle control exercised through threats of force or the withdrawal of family benefits and security. In these contexts, the rights of women (and girls) to control their own lives, to liberty or freedom of expression, association, movement and bodily integrity mean very little.
[76] Put shortly, “honour” is a form of rigid control of women by men. See the report by Dr Yakin Ertürk, Mission to Turkey (5 January 2007) at paras [19] to [21]:
[19] ... honour (Namus) is an important value in Turkish society; it serves to reproduce the rigid control exercised over women and their sexuality. For women, compliance with the code of honour can range from having to dress and act with “modesty”, to observing chastity prior to marriage, to accepting arranged marriages and not leaving the house without the consent and/or accompaniment of an older relative. In the region, honour is particularly important as the norm is codified into customary law, which is referred to as töre (custom or law). According to Töre, the family must ensure that the code of honour is observed by its members as transgressions (or mere rumours of such transgressions) are seen as “stains” on the entire family. These stains may have to be cleansed at any cost, if necessary through murder.
[20] Ties based on collective relationships reinforce and harden the namus ideology, as it is presented as part and parcel of the collective identity. Women and their immediate families are locked into the normative order of töre. Töre is all-encompassing: it regulates the moral order not only of women but also men - the safekeepers of that order - and the relationship among them. The observance of honour therefore becomes a societal concern.
[21] Tribal connections extend the reach of töre and make it harder for women to escape violence....
[77] As to whether a crime of “honour” may be distinguished from other crimes of violence against women, Dr Purna Sen in “‘Crimes of Honour’, Value and Meaning” in Lynn Welchman and Sara Hossain (eds), “Honour”: Crimes, Paradigms and Violence Against Women 42 at 48 points out that honour codes and crimes are not solely about individual men controlling the lives of individual women:
They are about community norms, social policing and collective decisions and acts of punishment....
[78] Her analysis is that while crimes of honour share a number of features with other forms of violence against women, they also have a number of characteristics that mark them out from other practices. The primary point of distinction is that a “crime of honour” is an action that removes from the collectivity [the family] the stain of dishonour. See op cit at 50:
In summary, crimes of honour are actions that remove from a collectivity the stain of dishonour, both gendered and locally defined, through the use of emotional, social or physical coercion over a person whose actual or imputed actions have brought that dishonour; physical force may involve killing the transgressor of the code of honour.
[79] At op cit 51 she goes on to observe that in cultures where codes of honour operate, “there is an overwhelming drive and motivation to collective morality, values and behaviours that conform with prevailing codes” and that one feature specific and particular about honour-based killings of women is “their being rooted in collectively monitored and policed codes of behaviour, the policing being in part carried out through killings”.
Conclusions on honour crimes
[80] Drawing on these converging analyses by Turkish and non-Turkish experts, we are persuaded that honour is an important value in Turkish society and that it enforces rigid control by men over women and their sexuality. It is about policing community norms and codes of behaviour, collective decisions and acts of punishment. Ultimately, it is about the distribution and exercise of power in Turkish society; or as phrased in the Ertürk report, it is about the rigid patriarchal oppression of women. In eastern and southeastern Turkey, honour as a norm has been codified into customary law and regulates the moral order not only of women but also the keepers of that order, namely men. The observance of honour is a societal concern and reflects the gendered inequality of power in that society.
[81] The question which next arises is whether it is possible for the political opinion ground of the Refugee Convention to be engaged by this context.
Political opinion - a gender-sensitive interpretation
[82] Being itself a definition, Article 1A(2) of the Convention does not define any of the five Convention grounds. Academic commentary suggests that “political opinion” should be understood “in the broad sense, to incorporate, within substantive limitations now developing generally in the field of human rights, any opinion on any matter in which the machinery of State, government, and policy may be engaged”: Guy S Goodwin-Gill and Jane McAdam, The Refugee in International Law, 3rd ed (2007, Oxford) at p 87. To similar effect see the background paper “Gender-Related Persecution” prepared for the September 2001 San Remo Global Consultations roundtable and published in Feller, Türk and Nicholson (eds), Refugee Protection in International Law: UNHCR’s Global Consultations on International Protection (2003, Cambridge) at 319, 346 and the UNHCR Guidelines on International Protection: Gender-Related Persecution Within the Context of Article 1A(2) of the 1951 Convention and/or its 1967 Protocol relating to the Status of Refugees (HCR/GIP/02/01, 7 May 2002) at para [32] (the UNHCR Guidelines on Gender-Related Persecution). Professor James C Hathaway in The Law of Refugee Status (1991, Butterworths) at 157 argues that the broad characterization of political opinion is an important means of maintaining the Convention’s vitality in circumstances where the basis for oppression may not be readily ascertainable.
[83] In giving the political opinion ground a gender-sensitive interpretation we take as our starting point the proposition that what is a political opinion is not a matter of definition but depends on the context of the case: Spijkerboer, Women and Refugee Status: Beyond the Public/Private Distinction (Emancipation Council, The Hague, 1994) p 58 and the UNHCR Guidelines on Gender-Related Persecution at para [32]. As Professor Spijkerboer explains at op cit 58:
Private talk in itself can be subversive, and therefore a political act, as in Orwell’s 1984. In the context of refugee law, cooking will normally be a private act, and therefore irrelevant. This may change, however, if the food is given to a political opponent of the authorities, or if the cooking is done communally by relatives of “disappeared” persons. There is political talk and private talk - as we know. But there is also private cooking and political cooking.
Therefore, an analysis of refugee law that uses the public/private distinction has to be on its guard. Public and private are not aspects of acts. They are aspects of analyses, be it by the authorities of the country of origin of an applicant (who may find cooking political), be it by an asylum adjudicator (who may find cooking inherently private)....
[84] Account must also be taken of how power is distributed and exercised in the particular society. The political opinion ground must be oriented to reflect the reality of women’s experiences and the way in which gender is constructed in the specific geographical, historical, political and socio-cultural context of the country of origin. In the particular context, a woman’s actual or implied assertion of her right to autonomy and the right to control her own life may be seen as a challenge to the unequal distribution of power in her society and the structures which underpin that inequality. In our view such situation is properly characterised as “political”. The point is succinctly encapsulated by the Refugee Women’s Legal Group (RWLG), Gender Guidelines for the Determination of Asylum Claims in the UK (July 1998) (reproduced in Crawley, Refugees and Gender: Law and Process, Appendix 3) at paras 4.13, 4.16 to 4.18:
4.13 It is important not to underestimate or overlook the political dimensions of a woman’s experiences of persecution even though a woman may not regard herself as making a political statement. She may not directly claim, orally or in writing, that she has been persecuted for reasons of political opinion and may find it difficult to explain the reasons for her persecution.
...
4.16 “Private” issues commonly associated with women are not inherently less political than those taking place in the “public” sphere. Conflicts concerning the demarcation of privacy (for example, freedom to choose to wear the veil or not, to have an education or undertake certain work, to be sexually active or not, to choose her partner, to be free from male domination and violence, to exercise reproductive rights and to reject female genital mutilation) are conflicts of a political nature.
4.17 A woman who opposes institutionalised discrimination against women or expresses views of independence from the social or cultural norms of society may sustain or fear harm because of her actual political opinion or a political opinion that has been or will be imputed to her. She is perceived within the established political/social structure as expressing politically antagonistic views through her actions or failure to act. If a woman resists gendered oppression, her resistance is political.
4.18 Where a woman does not directly or intentionally challenge institutionalised norms of behaviour she may nonetheless be imputed (ie attributed) with a political opinion. This can be seen, for example, in the characterisation of a raped woman as adulterous, in the social ostracism of an unmarried, separated, divorced, widowed or lesbian woman, and in the politicisation of (unintentional) violations of dress codes.
[85] Recognising these imperatives the Authority’s gender jurisprudence has for some years given a gendered interpretation to the political opinion ground. So in Refugee Appeal No. 2039/93 Re MN (12 February 1996), a claim based on a risk of being killed in an honour crime and on opposition to gendered oppression of women in Iran, the analysis centred on the power exercised both by men and by the state over women in that country. The challenge by the particular claimant to tribal values and to the religious ideology deployed to control women and to subject them to a wide range of discriminatory treatment was held to be properly characterised as within the political opinion ground of the refugee definition. That decision was followed and applied in Refugee Appeal No. 71427/99 (16 August 2000); [2000] NZAR 545; [2000] INLR 608 at paras [86] to [88], a case in which a successful claim to refugee status was based on domestic violence and systemic gender-based discrimination in Iran as highlighted by a child custody battle.
[86] It must be added that in orienting the “political opinion” ground to reflect the reality of women’s experiences, care must be taken to avoid the inadvertent favouring of the articulate claimant who is able to identify and explain the political nature of her opinion, actual or imputed and, where relevant, the political nature of her activities. A woman from the same country who has undergone the same experiences but who is not articulate and who may herself be unaware of the political context or construction of her views and actions must not be disadvantaged by her lack of awareness, insight or want of fluency. While the obligation to establish the refugee claim rests on the claimant, the inquiry is a shared one. It may become the responsibility of the decision-maker to identify that which in many gender-related cases is the unarticulated premise for the predicament of the claimant’s exposure to the risk of serious harm.
[87] Finally, in the refugee determination process it must be remembered that the construction of gender identity occurs in specific geographical, historical, political and socio-cultural contexts. It is these contexts which provide the answer to the question whether the risk of being persecuted is for reason of political opinion, not an abstract notion of how “political opinion” is to be defined. See further Heaven Crawley, “Gender, Persecution and the Concept of Politics in the Asylum Determination Process” (2000) 9 Forced Migration Review 17.
[88] With these observations as to how a gender-inclusive and gender-sensitive interpretation might be given to the political opinion ground we turn next to the determination whether on the facts the appellant is at risk of being persecuted for reasons of political opinion.
Political opinion - whether established on the facts
[89] A superficial analysis of the events of December 2006 (throwing wedding ring at the husband and telling him that the relationship was over) is that a “private” relationship had reached its point of termination. But in truth far more was involved. While the incident was geographically located in New Zealand, the husband and wife inhabit a very different political and socio-cultural world. They are Alevi Kurds originally from southeastern Turkey. For both, compliance with the code of honour (namus) is required by their families and by their broader community. Strict patriarchal control is exercised over the appellant. Töre (custom or law) requires her husband and each of the families to enforce that code of honour. The appellant’s severance of her relationship with her husband was an unambiguous act of self-emancipation from an abusive relationship and the structures of power and inequality which had sanctioned that relationship from the moment the appellant had been forced into it. Her unilateral action in ending the marriage can only be seen by the respective families as a direct challenge to her duties, to their power over her and to their own obligation under custom or law (töre) to police the collective code of honour by removing from the collectivity the stain of dishonour. The evidence shows that the death threats from the husband’s family followed immediately they learnt of the appellant’s actions.
[90] In the specific context we are satisfied that the appellant’s assertion of her right to life and of her right to control her life was a challenge to the collective morality, values, behaviours and codes of the two families and beyond them, of the greater “community” of which they are a part. This challenge to inequality and the structures of power which support it is plainly “political” as that term is used in the Refugee Convention. The appellant’s wish to be liberated from those structures is in this context a political opinion. It is for holding that opinion that she is at risk. Applying the causation standard discussed earlier, the appellant has established that she is at risk of being persecuted “for reasons of” political opinion.
[91] While on one view the holding that the claim to refugee status succeeds on the political opinion ground makes it unnecessary to examine the social group category, we will nevertheless explain briefly why the claim succeeds on that ground too.
[92] For the reasons explained in Refugee Appeal No. 71427/99 [2000] NZAR 545; [2000] INLR 608 at paras [92] to [105] it is indisputable that sex and gender can be the defining characteristic of a social group and that “women” may be a particular social group. See now also Fornah v Secretary of State for the Home Department [2007] 1 AC 412 (HL) where Lord Bingham at [31], Lord Hope at [54] and Baroness Hale at [114] acknowledged that on the facts the particular social group could be defined as women in Sierra Leone; just as in R v Immigration Appeal Tribunal; Ex parte Shah [1999] 2 AC 629 (HL) Lord Steyn at p 644 and Lord Hoffman at p 654 accepted that the evidence established that women in Pakistan were a particular social group and that there was no need to provide a more narrow definition. While the issue was left undecided in Minister for Immigration and Multicultural Affairs v Khawar (2002) 210 CLR 1 (HCA), Gleeson CJ at [32] to [35] and Kirby J at [126] and [127] were disposed to accepting that women in Pakistan are a particular social group.
Particular social group - whether established on the facts
[93] In the present case the expert evidence, particularly that of Dr Yakin Ertürk in her report of 5 January 2007, is that honour is an important value in Turkish society generally. While patriarchal power is particularly emphasised in eastern and southeastern Turkey, women generally are assigned an inferior status in Turkish society. See also the passages from the Department of State Country Reports on Human Rights Practices for 2007: Turkey (March 2008) referred to earlier. As Dr Ertürk explains in her report, despite modernisation, the gender order remains intact in Turkey. This opinion is shared by Dr Kogacioglu, as will be clear from the earlier discussion of the country information.
[94] Without re-sifting the country information earlier set out in this decision, we find that on the facts the appellant has established that she is a member of a particular social group, namely women in Turkey. It is because she is a member of that group that she is at risk of harm from non-state agents (her husband, his family and her own family) and the reason why effective state protection is not available to her. Applying the causation standard discussed earlier, it can properly be said that the risk of being persecuted faced by her is “for reasons of” her membership of the identified particular social group.
Introduction
[95] The appellant having established a well-founded fear of being persecuted for reasons of her political opinion and membership of a particular social group, it follows that she satisfies the inclusion criteria stipulated by Article 1A(2) of the Refugee Convention. There are no exclusion issues.
[96] It might be thought that in those circumstances recognition of the appellant’s status as a Convention refugee must follow as a matter of course. The more so given that refugee status is declaratory, not constitutive. As the UNHCR Handbook on Procedures and Criteria for Determining Refugee Status at para [28] states, a person is a refugee within the meaning of the Convention as soon as she fulfils the criteria contained in the definition. This would necessarily occur prior to the time at which her refugee status is formally determined. Recognition of her refugee status does not therefore make her a refugee but declares her to be one. She does not become a refugee because of recognition, but is recognised because she is a refugee.
[97] Yet States party to the Convention, including New Zealand, have used the internal protection alternative (IPA) (in other jurisdictions referred to as an internal flight alternative or internal relocation alternative) to withhold recognition of refugee status from persons at risk of being persecuted for a Convention reason where that risk exists in part, but not all, of the country of origin. The logic of this approach is explained by Professor James C Hathaway in his introduction to the Michigan Guidelines on the Internal Protection Alternative (1999) 21 Mich. J. Int’l L. 131:
International refugee law is designed only to provide a back-up source of protection to seriously at-risk persons. Its purpose is not to displace the primary rule that individuals should look to their state of nationality for protection, but simply to provide a safety net in the event a state fails to meet its basic protective responsibilities. As observed by the Supreme Court of Canada, “[t]he international community was meant to be a forum of second resort for the persecuted, a ‘surrogate’, approachable upon the failure of local protection. The rationale upon which international refugee law rests is not simply the need to give shelter to those persecuted by the state, but ... to provide refuge to those whose home state cannot or does not afford them protection from persecution.”
It follows logically that persons who face even egregious risks, but who can secure meaningful protection from their own government, are not eligible for Convention refugee status. Thus, courts in most countries have sensibly required asylum seekers to exhaust reasonable domestic protection possibilities before asserting their entitlement to refugee status. Where, for example, the risk of persecution stems from actions of a local authority or non-state entity (such as a paramilitary group, or vigilante gang) that can and will be effectively suppressed by the national government, there is no genuine risk of persecution, and hence no need for surrogate protection. [Citations omitted]
[98] However, State practice has been divergent. See James C Hathaway and Michelle Foster, “Internal Protection/Relocation/Flight Alternative as an Aspect of Refugee Status Determination” in Feller, Türk and Nicholson (eds), Refugee Protection in International Law: UNHCR’s Global Consultations on International Protection (2003, Cambridge) 357, 358 & 360. The primary reason for this divergence is an absence of consensus as to the proper textual foundation for the internal protection alternative.
[99] The law in New Zealand has been settled since Butler v Attorney-General [1999] NZAR 205 (CA) and the Authority’s subsequent decision in Refugee Appeal No. 71684/99 (29 October 1999); [2000] INLR 165. In these decisions the textual foundation for the internal protection alternative is located in the “protection” element of the refugee definition. However, it is necessary that account now be taken of the two relatively recent decisions of the House of Lords in Januzi v Secretary of State for the Home Department [2006] 2 AC 426 and AH (Sudan) v Secretary of State for the Home Department [2007] 3 WLR 832 as well as the two recent decisions of the High Court of Australia in SZATV v Minister for Immigration and Citizenship (2007) 237 ALR 634 and SZFDV v Minister for Immigration and Citizenship (2007) 237 ALR 660. In all four decisions the analysis is located in the “well-founded” element of the definition.
[100] The primary issues are:
(a) The textual foundation for the internal protection alternative.
(b) The test to be applied for determining the circumstances in which recognition of refugee status can properly be withheld without violating the text, object and purpose of the Refugee Convention.
[101] It is to be recalled that Article 1A(2) of the Convention relevantly defines a refugee as a person who:
... owing to well-founded fear of being persecuted ... is unable, or ... unwilling to avail himself of the protection of that country.... [Emphasis added]
Currently the division of opinion is as to whether justification for the withholding of recognition of refugee status from a person at risk of being persecuted in one part of a country but not in another is to be found in the “well-founded” element or in the words “protection of that country”. In short, the issue is whether the analysis is one of risk or one of protection. In the United Kingdom, Australia and the European Union the analysis is referenced to the former whereas in New Zealand and in Canada it is referenced to the latter. We begin with the analysis based on protection.
The analysis based on protection
[102] In the seminal decision of Canada (Attorney General) v Ward [1993] 2 SCR 688 the Supreme Court of Canada explicitly accepted that where a refugee claimant is shown to have access to true internal protection inside his or her own country, refugee status need not be recognised. This is because international refugee law is designed only to provide a back-up source of protection to seriously at-risk persons. The following quote is taken from p 709:
International refugee law was formulated to serve as a back-up to the protection one expects from the state of which an individual is a national. It was meant to come into play only in situations when that protection is unavailable, and then only in certain situations. The international community intended that persecuted individuals be required to approach their home state for protection before the responsibility of other states becomes engaged. For this reason, James Hathaway refers to the refugee scheme as ‘surrogate or substitute protection’, activated only upon failure of national protection; see The Law of Refugee Status (1991), at p. 135.
And at p 722:
It is clear that the lynch-pin of the analysis is the state’s inability to protect: it is a crucial element in determining whether the claimant’s fear is well-founded, and thereby the objective reasonableness of his or her unwillingness to seek the protection of his or her state of nationality.
[103] Case law in Canada both before and after Ward has explicitly recognised that the internal protection alternative analysis is located in the protection element of the definition. See for example Zalzali v Canada (Minister of Employment and Immigration) [1991] 3 FC 605 (FC:CA) at 614-615:
I know that in principle persecution in a given region will not be persecution within the meaning of the Convention if the Government of the country is capable of providing the necessary protection elsewhere in its territory....
[104] The relevant extracts from Thirunavukkarasu v Canada (Minister of Employment and Immigration) [1994] 1 FC 489 (FC:CA) are at 592-593:
It should first be emphasized that the notion of an internal flight alternative (IFA) is not a legal defence. Neither is it a legal doctrine. It merely is a convenient, short-hand way of describing a fact situation in which a person may be in danger of persecution in one part of a country but not in another. The idea of an internal flight alternative is “inherent” in the definition of a Convention refugee ... it is not something separate at all. That definition requires that the claimants have a well-founded fear of persecution which renders them unable or unwilling to return to their home country. If claimants are able to seek safe refuge within their own country, there is no basis for finding that they are unable or unwilling to avail themselves of the protection of that country.
And at p 599:
Thus, the objective standard of reasonableness which I have suggested for an IFA is the one that best conforms to the definition of Convention refugee. That definition requires claimants to be unable or unwilling by reason of fear of persecution to claim the protection of their home country in any part of that country. The pre-requisites of that definition can only be met if it is not reasonable for the claimant to seek and obtain safety from persecution elsewhere in the country.
[105] The history and development of the internal protection principle in New Zealand is fully set out in Refugee Appeal No. 71684/99 (29 October 1999); [2000] INLR 165 and only some of the salient points will be referred to. It is sufficient to note that within months of the Authority first sitting on 10 June 1991 it declined to employ the nomenclature of “internal flight alternative”, holding that the inquiry is one based on protection. As explained in Refugee Appeal No. 11/91 Re S (5 September 1991) at 19:
... to pose any question postulated on an “internal flight alternative” is to ask the wrong question. Rather, the question is one of protection and is to be approached fairly and squarely in terms of the refugee definition, namely whether the applicant [is unable or, owing to such fear, is unwilling to avail himself of the protection of that country].
The Authority also accepted that the following passage from Professor Hathaway’s The Law of Refugee Status at p 134 contained a fair summary of the internal protection principle:
The logic of the internal protection principle must, however, be recognized to flow from the absence of a need for asylum abroad. It should be restricted in its application to persons who can genuinely access domestic protection, and for whom the reality of protection is meaningful. In situations where, for example, financial, logistical, or other barriers prevent the claimant from reaching internal safety; where the quality of internal protection fails to meet basic norms of civil, political and socio-economic human rights; or where internal safety is otherwise illusory or unpredictable, state accountability for the harm is established and refugee status is appropriately recognized.
[106] In the later decision of Refugee Appeal No. 135/92 Re RS (18 June 1993) the Authority drew together its jurisprudence on what it then (erroneously) described as the relocation issue and affirmed that the question was one of protection, not of flight. The inquiry required two issues to be addressed:
(a) Could the individual genuinely access domestic protection which is meaningful? This issue was extracted from the passage cited above from Professor Hathaway’s The Law of Refugee Status at p 134;
(b) Was it reasonable, in all the circumstances, to expect the individual to relocate? This limb of the test was extracted from the UNHCR Handbook at para [91].
The requirements were cumulative. In other words, before an individual possessing a well-founded fear of being persecuted could be expected to relocate within the country of origin, it had to be possible to say both that meaningful domestic protection could be genuinely accessed by that person and also that in all the circumstances, it was reasonable for that individual to relocate.
[107] The Authority’s jurisprudence on the internal protection issue was considered by the Court of Appeal in Butler v Attorney-General [1999] NZAR 205 (CA) (Richardson P, Henry, Keith, Tipping and Williams JJ). As to the “genuine access to meaningful protection” issue, the Court at pp 216-217 explicitly accepted that the location of the analysis is in the protection element of the definition:
Central to the definition of “refugee” is the basic concept of protection - the protection accorded (or not) by the country of nationality or, for those who are stateless, the country of habitual residence. If there is a real chance that those countries will not provide protection, the world community is to provide surrogate protection either through other countries or through international bodies. So both paragraphs of art 1A(2) define refugees in part by reference to their ability or willingness to avail themselves of the protection of their country of nationality or of habitual residence.
[108] The Court of Appeal at p 217 cited with approval the relevant extracts from the decision of the Supreme Court of Canada in Ward where it is acknowledged that the refugee scheme is surrogate or substitute protection activated upon failure of national protection and that the lynch-pin is the state’s inability to protect.
[109] As to the “ reasonableness” limb of the Authority’s then formulation of issues, the Court of Appeal held at pp 217-218 that the various references to and tests for “reasonableness” (Handbook) or “undue harshness” (Thirunavukkarasu) must be seen in context, or against the backcloth that the issue is whether the claimant is entitled to the status of refugee. It is not a standalone test, authorising an unconfined inquiry into all the social, economic and political circumstances of the applicant including the circumstances of members of the family. Rather, the reasonableness test in the “relocation” context must be related to the primary obligation of the country of nationality to protect the claimant. The refugee inquiry “is narrowly focused on the persecution and protection of the particular claimant”. The reasonableness test is not free standing. It must be tied back to the refugee definition and to the Convention’s purpose of surrogate protection from the risk of being persecuted. The question is whether, having regard to that purpose, it is unreasonable to require refugee claimants to avail themselves of the available protection of their country of nationality. See pp 217-218:
The various references to and tests for “reasonableness” or “undue harshness” (a test stated by Linden JA in Thirunavukkarasu v Minister of Employment and Immigration) must be seen in context or, to borrow Brooke LJ’s metaphor, “against the backcloth that the issue is whether the claimant is entitled to the status of refugee”, R v Secretary of State for the Home Department, ex parte Robinson p 435. It is not a stand alone test, authorising an unconfined inquiry into all the social, economic and political circumstances of the application including the circumstances of members of the family. The test is for instance sharply different from the humanitarian tests provided for in the Immigration Act 1987, ss 63B and 105. It does not in particular range widely over the rights and interests in respect of the family. The refugee inquiry is narrowly focussed on the persecution and protection of the particular claimant. ...
Rather than being seen as free standing (as more recent decisions of the Authority appear to suggest), the reasonableness test must be related to the primary obligation of the country of nationality to protect the claimant. To repeat what Professor Hathaway said in the passage relating to relocation quoted earlier, meaningful national state protection which can be genuinely accessed requires provision of basic norms of civil, political and socio-economic rights. To the same effect Linden JA in the Canadian case cited above, (1993) 109 DLR (4th) 682, 688, stresses that it is not a matter of a claimant’s convenience or of the attractiveness of the place of relocation. More must be shown. The reasonableness element must be tied back to the definition of “refugee” set out in the Convention and to the Convention’s purposes of original protection or surrogate protection for the avoidance of persecution. The relocation element is inherent in the definition; it is not distinct. The question is whether, having regard to those purposes, it is unreasonable in a relocation case to require claimants to avail themselves of the available protection of the country of nationality.
[110] It will be seen that both in principle and in law the Authority is bound to approach the “internal flight” or “relocation” alternative as an issue of protection, not one of well-foundedness. Once a refugee claimant has established a well-founded fear of being persecuted for a Convention reason in relation to at least part of the country of origin, recognition of that person as a Convention refugee can only be withheld if that person can genuinely access domestic protection which is meaningful.
The analysis based on risk
[111] In Januzi v Secretary of State for the Home Department [2006] 2 AC 426 (HL) the textual foundation of the analysis was located in the well-founded element of the definition. The following passage from the opinion of Lord Bingham at [7] (Lords Nicholls, Hope, Carswell and Mance agreeing) makes this clear:
The Refugee Convention does not expressly address the situation at issue in these appeals where, within the country of his nationality, a person has a well-founded fear of persecution at place A, where he lived, but not at place B, where (it is said) he could reasonably be expected to relocate. But the situation may fairly be said to be covered by the causative condition to which reference has been made: for if a person is outside the country of his nationality because he has chosen to leave that country and seek asylum in a foreign country, rather than move to a place of relocation within his own country where he would have no well-founded fear of persecution, where the protection of his country would be available to him and where he could reasonably be expected to relocate, it can properly be said that he is not outside the country of his nationality owing to a well-founded fear of being persecuted for a Convention reason.
[112] Lord Carswell at [65], in agreeing with Lord Bingham, confirmed that the focus is on the well-founded element:
Accommodation of these principles has been attempted by two avenues, the choice of which may be critical, for it may lead to different results in individual cases. The first is by focussing on the well-founded nature of the fear of the applicant for asylum, the approach of the Court of Appeal at paras 23-24 of its judgment in E v Secretary of State for the Home Department [2004] QB 531, which your Lordships have approved. On this basis, where there is a safe place of relocation in the applicant’s country, he does not have a well-founded fear of persecution if he returns to that part of his country and therefore does not satisfy the definition of a refugee.
[113] The second “avenue” referred to by Lord Carswell is the location of the analysis in the protection element of the definition. He dismissed this approach by referring to the reasons given by Lord Bingham at paras [15] to [19] of his opinion. However, in those paragraphs Lord Bingham addresses not the location of the analysis, but the conditions which must exist in the proposed site of internal protection before it can be said that it is reasonable to expect the refugee claimant to seek refuge in another part of the country of origin. In doing so he fundamentally misunderstands the protection test as it is applied in New Zealand. But for present purposes it is sufficient to note that none of the opinions in Januzi explain with any clarity or cogency why, as a matter of treaty interpretation, the analysis properly belongs in the well-founded element of the definition.
[114] The EU Refugee Qualification Directive (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) addresses internal protection in Article 8:
Article 8
Internal Protection
1. As part of the assessment of the application for international protection, Member States may determine that an applicant is not in need of international protection if in a part of the country of origin there is no well-founded fear of being persecuted or no real risk of suffering serious harm and the applicant can reasonably be expected to stay in that part of the country.
2. In examining whether a part of the country of origin is in accordance with paragraph 1, Member States shall at the time of taking the decision on the application have regard to the general circumstances prevailing in that part of the country and to the personal circumstances of the applicant.
3. Paragraph 1 may apply notwithstanding technical obstacles to return to the country of origin.
[115] It will be seen that although the heading to Article 8 is “Internal Protection”, the text of the article locates the analysis in the well-founded element of the Convention definition. The reason for so doing is not revealed. But it is to be noted that the Qualification Directive is not “superior” to the Refugee Convention. On the contrary, it leaves the Convention intact as the governing international treaty on refugee status. See Hugo Storey, “EU Refugee Qualification Directive: A Brave New World? (2008) 20 IJRL 1, 7. The responsibilities of a State party to a refugee claimant flow from the Refugee Convention, not from the Qualification Directive. We do not accept that the Qualification Directive properly reflects the terms of the Convention in their context and in the light of its object and purpose. We also note that Article 8 has been challenged by the UNHCR itself, a point to which reference will be made shortly.
[116] In Australia the analysis is also located in the well-founded element of the refugee definition, though for very different reasons. In contrast to the Januzi decision where the textual foundation of the analysis is more asserted than demonstrated, the High Court of Australia in at least two decisions has explicitly held that “the protection of that country” in the refugee definition means “diplomatic or consular protection”, not protection from violation of fundamental human rights. It follows that the only place left for the “relocation” analysis is the “well-founded” element of the definition. In other words, the Australian position is a “nowhere else to go” position based (as will be shown) on a misinterpretation of the “protection” element of the definition.
Critique of the analysis based on risk
[117] The question whether locating the analysis in the well-founded fear element is justifiable as a matter of international law is addressed in the San Remo paper by James C Hathaway and Michelle Foster, “Internal Protection/Relocation/Flight Alternative as an Aspect of Refugee Status Determination” in Feller, Türk & Nicholson (eds), Refugee Protection in International Law: UNHCR’s Global Consultations on International Protection (2003, Cambridge) at p 357. The authors acknowledge that the elements of “well-founded fear” and “protection” are to some extent intertwined. If the State can effectively suppress the risk of serious harm in the claimant’s place of origin, then the person does not have a well-founded fear of being persecuted. However, they draw attention, to the legal fiction inherent in locating the analysis in the well-founded element. They point out that an inquiry into an internal flight alternative or into relocation cannot logically arise unless and until the refugee claimant has established that in at least part of the country of origin there is a well-founded fear of being persecuted for a Convention reason. It is only if such well-founded fear is established in relation to some part of the country that the issue arises as to whether the person should find an internal flight alternative or, as it is referred to in Januzi, to relocate. The internal flight alternative or internal relocation alternative is in truth being used to deny refugee status to a person who satisfies the inclusion clause criteria stipulated by Article 1A(2) of the Convention in part, but not all, of the country of origin. The point is succinctly explained at 366-368:
It is crucial to understand, however, that the analysis shifts significantly once it has already been established that a person has well-founded fear of being persecuted in his or her home area (region ‘A’), which of course implies that the State is unable or unwilling to protect the person in that region. Once this is established, it is neither logical nor realistic to find that the fact that the State can protect the person in some other region of the country (region ‘B’) means that she no longer has a well-founded fear of being persecuted in region A. The well-founded fear of being persecuted in region A has not been negated or removed by the provision of national protection in region B, just as the risk would not be removed or negated by the availability of protection in a country of second nationality or in an asylum State. In all of these cases, the refugee continues to face a well-founded fear of being persecuted in region A of his or her country of origin, but is able to avail him or herself of countervailing national protection. To hold otherwise is to construct a legal fiction fundamentally at odds with common sense.
Indeed, the text of the 1951 Convention itself envisages that the possibility of national protection will not necessarily allay the well-founded fear, as was well explained by Sedley LJ in the seminal Karanakaran decision:
[B]oth the special adjudicator and the tribunal failed to approach the Convention methodically. They treated the availability of internal [protection] as a reason for holding that the fear of persecution was not well-founded. There may possibly be countries where a fear of persecution, albeit genuine, can so readily be allayed in a particular case by moving to another part of the country that it can be said that the fear is either non-existent or not well-founded, or that it is not “owing to” the fear that the applicant is here. But a clear limit is placed on this means of negating an asylum claim by the subsequent provision of the Article that the asylum-seeker must be, if not unable, then unwilling because of ‘such fear’ - ex hypothesi his well-founded fear of persecution - to avail himself of his home state’s protection. If the simple availability of protection in some part of the home state destroyed the foundation of the fear or its causative effect, this provision would never be reached.
[citations omitted]
[118] One can add to the Hathaway and Foster “legal fiction” point the acknowledgment of Kirby J in SZATV v Minister for Immigration and Citizenship (2007) 237 ALR 634 (HCA) at [63] that locating the inquiry in well-foundedness strains the language of the Convention and justifiably attracts the cogent criticism that approaching the problem in this way “involves building an edifice of reasoning on a very scant textual foundation”.
[119] Hathaway and Foster at 368-370 go on to identify the negative practical consequences of anchoring the internal flight or relocation analysis in the well-founded element of the definition:
(a) First, it has led some States and courts to assert a requirement that the applicant establish “country-wide persecution”. As they point out at 369, this approach is not justified by the text of the Convention. Rather it requires additional restrictive words to be read into the Article 1A(2) definition such that it reads: well-founded fear of being persecuted throughout the country of nationality. Moreover, it imposes an extremely onerous burden on refugee applicants and tends to produce “a wide-ranging fishing expedition into potential alternative protection regions, risks ‘the conflation of issues’ and a ‘consequent lack of focused analysis’”. The authors rightly observe that by contrast, analysing internal flight or relocation as a protection alternative provides structure to the determination exercise and encourages a logical, methodical approach to the determination process. It is thus of considerable assistance to decision-makers as well as to applicants (p 370):
A protection-based understanding of IFA reinforces the fact that once the applicant has established a well-founded fear in one location, she is entitled to the full weight of the establishment of a prima facie case. In this way, the IFA analysis is understood as akin to an exclusion enquiry such that the evidentiary burden is then on the party asserting an IFA to establish that it exists.
(b) A second major practical concern is that conceiving the internal flight issue as part of the initial inquiry into the existence of a well-founded fear of being persecuted encourages decision-makers to pre-empt the analysis of well-founded fear in the first region by moving directly to the question of an internal flight alternative. There is a real danger of the claim being dismissed without consideration of the conditions giving rise to the well-founded fear in the region of origin. At p 371 the authors observe:
... by commencing the enquiry with an assessment of the well-founded fear in the region from which the person fled before moving to the protection question, the decision maker has a clear benchmark against which to assess the sufficiency of the internal protection available to the applicant. To locate the analysis within the well-founded fear criterion, on the other hand, allows the decision maker to avoid this careful analysis, and raises a substantial risk that legitimate claims will be dismissed following only cursory consideration of the relevant circumstances.
For recent illustration of this recurring error see UNHCR, Asylum in the European Union: A Study of the Implementation of the Qualification Directive (November 2007) at 58.
[120] For the reasons given by Hathaway and Foster we remain of the view first expressed in Refugee Appeal No. 11/91 Re S (5 September 1991) at 19 and affirmed in an unbroken line of Authority decisions through to Refugee Appeal No. 71684/99 (29 October 1999); [2000] INLR 165 that as a matter of treaty interpretation the proper location of the inquiry is in the protection element of the Refugee Convention. That is why, since September 1991 the Authority has framed the analysis as a protection inquiry, not one of flight or relocation. Happily, as a matter of law, it is bound by the decision in Butler v Attorney-General [1999] NZAR 205 (CA) to take this view. But even were the Authority free to do so, it would not depart from its earlier jurisprudence. The reasons given by Hathaway and Foster for locating the analysis in the protection element of the refugee definition are compelling, as Kirby J acknowledged in SZATV. With the greatest of respect, we find nothing persuasive to the contrary in Januzi and AH (Sudan).
Protection as diplomatic protection
[121] For a statement of the view that protection means “diplomatic protection” see Minister for Immigration and Multicultural Affairs v Khawar (2002) 210 CLR 1 (HCA) at [62] per McHugh and Gummow JJ:
The definition of “refugee” is couched in the present tense and the text indicates that the position of the putative refugee is to be considered upon the footing that that person is outside the country of nationality. The reference then made in the text to “protection” is to “external” protection by the country of nationality, for example by the provision of diplomatic or consular protection, and not to the provision of “internal” protection provided inside the country of nationality from which the refugee has departed.
[122] To similar effect see Minister for Immigration and Multicultural Affairs v Respondents S152/2003 (2004) 222 CLR 1 (HCA) at [19] (the joint judgment of Gleeson CJ, Hayne and Heydon JJ):
The immediate context is that of a putative refugee, who is outside the country of his nationality and who is unable or, owing to fear of persecution, unwilling to avail himself of the protection of that country. As explained in Khawar, we accept that the term “protection” there refers to the diplomatic or consular protection extended abroad by a country to its nationals. In the present case, the first respondent must show that he is unable or, owing to his fear of persecution in Ukraine, unwilling to avail himself of the diplomatic or consular protection extended abroad by the State of Ukraine to its nationals. Availing himself of that protection might result in his being returned to Ukraine. Where diplomatic or consular protection is available, a person such as the first respondent must show, not merely that he is unwilling to avail himself of such protection, but that his unwillingness is owing to his fear of persecution. He must justify, not merely assert, his unwillingness.
[123] This reading of “protection” has been applied by the High Court of Australia in the two recent cases referred to earlier. As in the House of Lords, the terminology preferred in these cases is “relocation”. See SZATV v Minister for Immigration and Citizenship (2007) 237 ALR 634 per Gummow, Hayne & Crennan JJ at [16] & [17] with Callinan J at [105] agreeing and SZFDV v Minister for Immigration and Citizenship (2007) 237 ALR 660 where the same four Judges explicitly followed Januzi by locating the analysis in the well-founded element.
[124] In both decisions Kirby J gave separate but concurring judgments. It is notable, however, that when addressing the textual foundation for the “relocation” test, Kirby J in SZATV at [60] and [61] expressed the view that the High Court should reconsider its holding that “protection” means “diplomatic protection”. He observed that the contrary interpretation (that the analysis belongs in the protection element of the definition) appeared more persuasive and more relevant to the central purposes of the Refugee Convention. He nevertheless felt bound to apply the “diplomatic protection” meaning of “protection” until the High Court itself reconsidered the issue. Given the condition of the law in Australia, he conceded that the only possible textual basis left to afford a foundation for the suggested relocation test was thus “the notion of ‘well-foundedness’”.
[125] In casting doubt on the “diplomatic protection” interpretation, Kirby J drew heavily on the Global Consultations paper by James C Hathaway and Michelle Foster referred to earlier.
[126] The “diplomatic protection” interpretation appears to be unique to Australia and is not shared by other Commonwealth or European Union countries. It was an approach explicitly rejected by Lord Carswell in Januzi at [66].
Critique of the “diplomatic protection” interpretation
[127] Hathaway and Foster at op cit 373 to 380 identify significant problems with the “diplomatic protection” analysis:
(a) First and most obviously, the extended term “diplomatic protection” does not appear in the text of the Convention itself and the ways in which the term “protection” is used elsewhere in the Convention make the High Court of Australia position anomalous. In particular, the Preamble refers to the intention of the parties to “revise and consolidate previous international agreements relating to the status of refugees and to extend the scope of and the protection afforded by such instruments”, and to the importance of coordinated measures to facilitate UNHCR’s task of “supervising international conventions providing for the protection of refugees”. The authors point out that clearly, “protection” as referred to in the Preamble cannot mean only “diplomatic protection”, since the Convention is concerned nearly exclusively with the provision of “protection” understood in the sense of human rights protection.
(b) Second, the isolated historical references that can be located to diplomatic protection do not justify the conclusion that the framers of the 1951 Convention clearly had this highly specialised understanding of “protection” in mind. There is simply too little historical evidence to justify the conclusion that the authors of the Convention specifically assigned to the term “protection” the special meaning of “diplomatic protection”.
(c) Third, even if it were somehow shown that the special meaning of “diplomatic protection” should inform the Convention’s general references to “protection”, the view as espoused by the High Court of Australia misinterprets the notion of diplomatic protection under international law. It has a narrow and well-established meaning as “action taken by a State against another State in respect of an injury to the person or property of a national caused by an internationally wrongful act or omission attributable to the latter State”. This is a considerably narrower concept than that advanced by proponents of the “protection equals diplomatic protection” view in the refugee context. Thus, not only do proponents of this view seek to read additional words into the Convention text, they also substitute the precise and well-established understanding of the term created by the addition of these words with a modified and expanded version of this term of art in international law. As Hathaway and Foster point out, this analysis simply cannot be maintained as a matter of treaty interpretation. It should be added that the narrow reading of diplomatic protection has been affirmed recently by the International Court of Justice in Ahmadou Sadio Diallo (Republic of Guinea v Democratic Republic of the Congo) - Preliminary Objections (24 May 2007) (<http://www.icj-cij.org>) at para [39] citing with approval the summary of customary international law in Article 1 of the draft Articles on Diplomatic Protection of the International Law Commission:
diplomatic protection consists of the invocation by a State, through diplomatic action or other means of peaceful settlement, of the responsibility of another State for an injury caused by an internationally wrongful act of that State to a natural or legal person that is a national of the former State with a view to the implementation of such responsibility.
See also generally Chittharanjan F Amerasinghe, Diplomatic Protection (2008, Oxford) Chapters 3 and 9.
(d) Fourth, it is difficult to justify the “diplomatic protection” interpretation as being consistent with the object and purpose of the 1951 Convention (a primary requirement of treaty interpretation) since its relevance to refugee status is not clear. In addition, the decision whether or not to exercise diplomatic protection may reflect considerations of domestic or international politics that have no relationship to the ability of a State to protect its nationals internally. Thus, arguing that refugee status should turn upon the willingness or ability of the State to exercise diplomatic protection is illogical.
(e) Fifth, the availability of diplomatic protection does not necessarily bear any relationship whatsoever to the question whether a State would wish to protect an individual against a well-founded fear of being persecuted. The diplomatic protection thesis allows the unilateral action of the State of nationality to remove the refugee’s right to protection, a position irreconcilable with Article 1C(1) which denies status only where the refugee voluntarily re-avails him or herself of the protection of the State of nationality.
(f) Sixth, the Vienna Convention on the Law of Treaties, 1969 specifically directs attention to subsequent practice in the application of a treaty which establishes agreement of the parties regarding its interpretation. The leading cases in common law jurisdictions reveal that the narrow “diplomatic protection” approach is inconsistent with a growing body of jurisprudence that recognises that the object of the Convention is to afford protection to those for whom neither is available in their own country.
(g) Seventh, the assertion that “protection” should be understood as “diplomatic protection” is also out of step with most contemporary pronouncements of UNHCR.
[128] The authors conclude at p 380 that there is no compelling reason to force a narrow, decontextualized reading of “protection” onto the Convention.
[129] It was to this conclusion that Kirby J referred in SZATV at [58] to [63] when he expressed the opinion that:
Overseas courts have not followed this court’s view of the meaning of “protection” in this context. In my view this court should reconsider its holding in this respect. The contrary view appears more persuasive. Moreover, it is one more relevant to the central purposes of the Refugees Convention. It is also more relevant to the issue under consideration in this appeal.
Conclusion as to textual foundation for the IPA
[130] Both in principle and in law the Authority is bound to approach the “internal flight or relocation alternative” as an issue of protection, not one of well-foundedness. Once a refugee claimant has established a well-founded fear of being persecuted for a Convention reason, recognition of that person as a Convention refugee can only be withheld if that person can genuinely access domestic protection which is meaningful. There are four elements to this inquiry: an ability to access the proposed internal protection alternative, the absence of a risk of being persecuted in that place, the absence there of other forms of serious harm and the ability to access a minimum level of civil, political and socio-economic rights.
[131] Unfortunately, there is no international consensus on these four points. Just as there is divergence as to the location of the internal protection alternative analysis, there is divergence also as to the test for withholding recognition of refugee status. It is this issue which is addressed next.
The “reasonableness” test
[132] While the courts in Canada and the United Kingdom diverge as to the location of the “flight” or “relocation” analysis, they do apply a common test for withholding recognition of refugee status. In shorthand, this is known as the reasonableness test, namely whether it is reasonable to expect the refugee to seek his or her safety in a different part of the country of origin or, to put the issue another way, whether it would be unduly harsh to expect the person to relocate within the country of origin before seeking refugee status abroad. See Thirunavukkarasu at 592-593, cited with approval by Lord Bingham in Januzi at [12].
[133] This is not, however, the test that has been applied in New Zealand post-Butler.
The reasonableness test in New Zealand after Butler
[134] For the reasons explained in Refugee Appeal No. 71684/99 at paras [51] to [73] the “reasonableness” limb of the protection inquiry was abandoned by the Authority in October 1999 following the decision of the Court of Appeal in Butler that there is no freestanding “reasonableness” test and that the analysis must be tied back to the refugee definition and to the Convention’s purpose of surrogate protection from the risk of being persecuted. Since then the analysis has not employed language and concepts (“flight”, “relocation” and “reasonableness”) extraneous to the Convention. Rather, the inquiry has been anchored in the refugee definition itself, particularly in the “protection” element. Moving away from a test based on “reasonableness”, the Authority has quantified “meaningful state protection” by identifying the preconditions which must exist before a finding can be made that an internal protection alternative is available. Of the original pre-Butler test (protection plus reasonableness), only the first limb remains. The decision-maker must decide whether the individual can genuinely access domestic protection which is meaningful. Informed by the Michigan Guidelines on the Internal Protection Alternative (1999) 21 Mich. J. Int’l L. 131, the revised approach identifies four prerequisites before it can be said that genuine access to meaningful state protection is established. First, the proposed site of protection must be accessible to the individual, meaning access which is practical, safe and legal. Second, in the proposed site of internal protection, the real chance of being persecuted for a Convention reason must be eliminated. In the language of Hathaway and Foster at op cit 389, the internal protection alternative must offer an “antidote” to the well-founded fear of being persecuted shown to exist in the applicant’s place of origin. Third, in the proposed site of internal protection there must be no new risks of being persecuted or of being exposed to other forms of serious harm or of refoulement to the region of origin. Fourth, the individual must be able to access in the proposed internal protection alternative a minimum standard of affirmative State protection, that standard being measured against the standard of protection prescribed by the Refugee Convention itself.
[135] If these preconditions are satisfied it necessarily follows that it would be “reasonable” to expect the claimant to avail him or herself of that protection. But rather than posing the test in terms of a conclusion (whether it is “reasonable”), the Authority has formulated a structured test which is tied to the specific language of Article 1A(2), and which stipulates objective criteria for investigation by the decision-maker. By eschewing a test based on subjective notions of “reasonableness”, it also precludes the intrusion of non-protection related issues into the decision-making process, being the point made in the Butler decision.
Criticisms of the reasonableness test
[136] As pointed out by Hathaway and Foster at op cit 383 and 385, the term “reasonableness” is vague and open to vastly divergent subjective interpretation. In practice, the reasonableness formulation allows decision-makers to assess the refugee claimant’s alternatives in light of their own view of what constitutes “reasonable” behaviour. It is “an amorphous test ... not amenable to structure or guidance by appellate courts”, a comment which perhaps aptly captures the experience of the courts in the Januzi and AH (Sudan) cases. Hathaway and Foster also note that the “reasonableness” test produces an “eclectic or ad hoc jurisprudence concerning claimants from the same countries and in similar situations” and this inherent lack of analytical clarity produces wide inconsistency between jurisdictions. They conclude at 388 that:
... it is this very lack of justification for the reasonableness test in the Convention text that makes it an unwieldy basis upon which to anchor the assessment of an internal alternative to asylum.
In their view at op cit 389, the textual centrality of the protection element of the definition:
... has the distinct advantage of being a standard actually derived from a treaty that States have formally agreed to be binding on them. In addition, it provides a focussed and principled framework of analysis, based on the aims and objects of the 1951 Convention.
[137] Subsequent experience in Europe has borne out these concerns. The inherent lack of analytical clarity to the reasonableness test has produced wide inconsistency between jurisdictions within the European Union, notwithstanding the uniformity intended by Article 8 of the Qualification Directive, cited with apparent approval in Januzi at [17]. The deadline for EU Member States to comply with the Qualification Directive was 10 October 2006. In the period between March and July 2007 the UNHCR engaged in a study of the implementation of the Qualification Directive in five member states and focused on six key issues, including that of internal protection (Article 8). The results of the study have been published by the UNHCR as Asylum in the European Union: A Study of the Implementation of the Qualification Directive (November 2007). In relation to the internal protection alternative (Article 8) it found that the extent to which the concept is applied to deny international protection varies and that the Qualification Directive is not yet achieving its aim to introduce a common concept of internal protection:
The treatment of Chechen applicants is just one example that highlights the divergence. Of the decisions reviewed in France, none applied the concept of internal protection to Chechens. In contrast, in Germany and the Slovak Republic, most parts of the Russian Federation are accepted as possible ‘internal protection alternatives’. The divergence in approach appears to be due in part to vastly differing interpretations of what is ‘reasonable’ under the Directive. These interpretations are not always in line with UNHCR guidance on this issue. [see p 11]
... there appears to be wide divergence across and within jurisdictions with regard to what constitute “reasonable stay”. [see p 60]
[138] Notwithstanding identification of this predictable flaw, the UNHCR’s own Guidelines on International Protection: “Internal Flight or Relocation Alternative” Within the Context of Article 1A(2) of the 1951 Convention and/or 1967 Protocol relating to the Status of Refugees (23 July 2003) continue to employ the “reasonableness” test. Being insufficiently tied back to the definition of “refugee” and to the element of protection, the Guidelines contain the same deficiencies identified and criticised by the New Zealand Court of Appeal in Butler. In the circumstances we do not find them (the Guidelines) persuasive. Furthermore, while the Guidelines usefully stress the now well-established point that there is a strong presumption against an internal protection alternative where the agent of persecution is the state, they give insufficient emphasis to the imperative that the internal protection alternative offer an “antidote” to the well-founded fear of being persecuted shown to exist in the applicant’s place of origin. In addition, insufficient emphasis is given to the requirement that the proposed site of internal protection contain no new risks of being persecuted or of exposure to other serious harms. While the Guidelines do require that in the proposed area there be respect for basic human rights standards and socio-economic conditions that will allow, for example, the claimant to earn a living or to access accommodation, this inquiry is regrettably framed in unhelpful language, namely whether the individual can lead a “relatively normal life without facing undue hardship”. There is insufficient emphasis on rights. As the more recent UNHCR Handbook for the Protection of Women and Girls (January 2008) notes at para [4.1.2], protection is, fundamentally, about rights. The test applied by the Authority gives recognition to this imperative by requiring the decision-maker to focus on rights as opposed to subjective understandings of “reasonableness”, “hardship” and “a normal life”.
The protection inquiry - identifying basic norms of civil, political and socio-economic rights
[139] In New Zealand “meaningful national state protection which can be genuinely accessed” requires the provision of basic norms of civil, political and socio-economic rights: Butler at p 218. In identifying what those basic norms are, the approach taken by the Authority, informed by the Michigan Guidelines on the Internal Protection Alternative, has been to acknowledge:
(a) That the issue of the internal protection alternative cannot logically arise unless and until the refugee claimant has established that in at least one part of the country of origin he or she has a well-founded fear of being persecuted for a Convention reason. This means that the individual has established, at least putatively, that he or she is a Convention refugee and the internal protection alternative inquiry is in truth an inquiry as to whether recognition of the person as a refugee is to be withheld.
(b) That if the claimant’s status as a refugee were to be recognised, he or she would be entitled to the rights found in the text of the Refugee Convention as set out in Articles 2 to 33 including, for example, freedom of religion, freedom of movement, access to courts, and right to work, social assistance and primary education.
[140] Because Articles 2 to 33 of the Convention set a minimum level of rights relating to what Hathaway and Foster at op cit 408 describe as “a core subset of civil and socio-economic rights” and because the internal protection alternative is an “exclusion” enquiry, recognition of refugee status can only be properly withheld where the refugee claimant can access in his or her own country the rights under the Refugee Convention which he or she would have been entitled to in the country of asylum. But the conditions in the proposed site of internal protection do not have to meet the requirements of international human rights in full. Most certainly the civil and socio-economic rights are not to be measured against those enjoyed by the citizens of the asylum state (here, New Zealand). Nor are refugee decision-makers expected to slavishly apply the Articles 2 to 33 standards. Rather, as Hathaway and Foster at op cit 405 to 409 make clear, decision-makers are to take inspiration from the interests protected by Articles 2 to 33 as a way of defining affirmative protection in the refugee context when deciding whether recognition of refugee status is to be withheld. Those interests include freedom of religion (Article 4), rights to property and work (Articles 13-15, 17-19), social welfare assistance (Articles 23 & 24), access to housing and education (Articles 21 & 22), access to the courts (Article 16) and mobility rights, including identity papers and travel documents (Articles 26-28). These rights are now explained and discussed by Professor Hathaway in his comprehensive The Rights of Refugees Under International Law (2005, Cambridge). As reference to these interests is made for the purpose of the internal protection alternative analysis, it is not relevant in this context to engage with the levels of attachment at which these rights would otherwise operate or with the stipulated comparator groups.
The challenge to the human rights inquiry in Januzi
[141] The inclusion of basic norms of civil, political and socio-economic rights in the internal protection inquiry has been challenged in Januzi even though several of their Lordships tacitly agreed that such basic norms cannot be excluded from what they call the “relocation” inquiry. In the discussion which follows the term “relocation” refers to the UK analysis, “protection” to the New Zealand approach.
[142] Their Lordships’ primary objection to what Lord Bingham in Januzi at [15] refers to as the “Hathaway/New Zealand rule” is that the rights entitlements in Articles 2 to 33 are rights of refugees in the country of asylum, not in the country of origin. See the opinion of Lord Bingham generally and Lord Hope at [46]. On a literal reading of the Convention that is clearly correct. But their Lordships’ observation misses the point. The “Hathaway/New Zealand” approach has regard to the rights regime in the Convention not as a prescription, but as a guide. The point is made by Hathaway and Foster at op cit 409:
It is important to understand, however, that the IPA approach embraced by the “Michigan Guidelines” does not suggest a literal application of Articles 2-33 in considering internal protection, but rather that decision-makers seek inspiration from the kind of interests protected by these Articles as a way of defining an endogenous notion of affirmative protection in the refugee context. While in some ways falling somewhat short of the standard of “protection” that would follow from assessment by reference to all key human rights or even to the “Guiding Principles on Internal Displacement”, the non-discrimination approached embodied by the 1951 Convention nonetheless provides a legally solid and contextualized assurance of durable protection.
[143] And later at op cit 409:
As this example makes clear, our point is not that the “reasonableness” approach cannot generate positive protection results for asylum seekers whose cases are subject to internal protection analysis. To the contrary, in the hands of experienced and thoughtful decision-makers, we believe the results will be largely the same. The difference, however, is that the greater analytical structure of IPA analysis and its more solid footing in international refugee law allow it more dependably to generate rights-regarding determinations of the reality of internal protection. By focusing on the provision of fundamental civil and socio-economic rights on a non-discriminatory basis - whether by reference to the whole spectrum of international human rights law, the “Guiding Principles on Internal Displacement”, or to the rights in the 1951 Convention itself - an understanding of “protection” that is readily amenable to appellate and other accountability is established. [Emphasis in original]
[144] The focus throughout the internal protection alternative is on the level of protection accessible in the country of origin. See Refugee Appeal No. 71684/99 (29 October 1999); [2000] INLR 165 at [61]:
There is considerable force to the logic that that putative refugee status should only be lost if the individual can access in his or her own country of origin the same level of protection that he or she would be entitled to under the Refugee Convention in one of the state parties to the Convention ... [Emphasis added]
[145] To similar effect see the Michigan Guidelines on the Internal Protection Alternative:
[22] At a minimum, therefore, conditions in the proposed site of internal protection ought to satisfy the affirmative, yet relative, standards set by this textually explicit definition of the content of protection. The relevant measure is the treatment of other persons in the proposed site of internal protection, not in the putative asylum country. [Emphasis added]
The “Hathaway/New Zealand rule” has been seriously misread by the House of Lords. As an aside, we mention that it is not accepted that the internal protection alternative test applied by the Authority is appropriately described or confined as the “Hathaway/New Zealand rule”. Rather, it is the test properly arrived at by the application of the principles of treaty interpretation in the Vienna Convention on the Law of Treaties, 1969. Nevertheless, for convenience, we have deployed their Lordships’ shorthand expression when analysing their decisions.
[146] A further misinterpretation of the “Hathaway/New Zealand rule” is the assertion that it requires the conditions in the proposed site of internal protection to meet “the requirements of international human rights law in full”: HGMO (Relocation to Khartoum) Sudan CG [2006] UKAIT 00062 (3 August 2006) at para [150]. It is difficult to understand how this criticism can be made. The Authority’s decision in Refugee Appeal No. 71684/99 at paras [57] to [60] explicitly rejects the proposition that the conditions in the proposed site of internal protection must meet the requirements of international human rights law in full and while the Michigan Guidelines on the Internal Protection Alternative at [21] recognise that good reasons may be advanced to refer “to a range of widely recognised international human rights in defining the irreducible core content of affirmative protection in the proposed site of internal protection”, they recommend at para [22] that as a minimum, the point of reference be the affirmative standards set by the Convention itself. Hathaway and Foster at op cit 405 to 409 equally do not argue for an absolute human rights standard, preferring instead that decision-makers take inspiration from the categories of interests protected by Articles 2 to 33 of the Convention as a way of defining affirmative protection in the refugee context. The “Hathaway/New Zealand” approach being clear, we are at a loss to understand why it has been so comprehensively mischaracterised in English jurisprudence.
[147] Lord Bingham’s point at [15] in Januzi is that there is nothing in any Article of the Convention from which the “Hathaway/New Zealand rule” (that in the proposed site of protection the claimant be able to enjoy a minimum standard of civil, political and socio-economic rights) may by any process of interpretation be derived. With respect, that is not so. The withholding of recognition of refugee status on the basis of an internal protection alternative has “suspect origins” (Hathaway and Foster op cit 361) or a “fragile footing in the text of the ... Convention” (Kirby J in SZATV v Minister for Immigration and Citizenship at [39]). It is therefore important that such principles as are applied to withhold recognition of refugee status from a person who satisfies the refugee definition are at least consistent with the language, object and purpose of the Convention. See Article 31(1) of the Vienna Convention on the Law of Treaties, 1969. As the Refugee Convention itself defines genuine access to international surrogate protection as encapsulating, at a minimum, those rights contained in Articles 2 to 33, it is consistent with both principle and the logic of the Convention for those same rights to be used to measure genuine access to meaningful state protection internally in the country of origin when withholding recognition of refugee status from a person who is presumptively a refugee. The assertion that the “Hathaway/New Zealand rule” cannot be derived by “any process of interpretation” is consequently misconceived.
[148] The Authority’s process of interpretation has been clearly articulated in Refugee Appeal No. 71684/99 at paras [57] to [62]. We repeat here only paras [61] and [62]:
[61] In essence, our reasoning is as follows. Because under New Zealand law the issue of internal protection does not arise unless and until a determination is made that the refugee claimant holds a well-founded fear of persecution for a Convention reason, the inquiry into internal protection is really an inquiry into whether a person who satisfies the Refugee Convention and who is prima facie a refugee - at least in relation to an identified part of the country of origin - should lose that status by the application of the internal protection principle. There is considerable force to the logic that that putative refugee status should only be lost if the individual can access in his or her own country of origin the same level of protection that he or she would be entitled to under the Refugee Convention in one of the state parties to the Convention. Clearly some state parties will accord to refugees a greater range of human rights and freedoms than the minimal standards prescribed by the Refugee Convention. Other states will barely be able to satisfy the Convention standards. But the Refugee Convention itself sets the minimum standard of human rights which the international community has agreed should be accorded to individuals who meet the Refugee Convention. The “loss” of refugee status by the application of the internal protection principle should only occur where, in the site of the internal protection, this minimum standard is met.
[62] The added attraction of this approach is that it provides decision-makers with an identified, quantified and standard set of rights common to all state parties, thereby facilitating consistent and fair decision-making. The three limbs of the proposed test also satisfy the requirement of the Court of Appeal in Butler that the assessment of the protection issue be tied back to the Article 1A(2) definition of refugee and to the Convention’s purposes of original protection. If on this analysis, there is no protection-related purpose for the reasonableness inquiry, it has no place in the determination of the internal protection alternative....
[149] Other reasons given by Lord Bingham in Januzi at paras [17] to [19] as to why the “Hathaway/New Zealand rule” should not be followed include:
1. That the “rule” is not expressed in the Qualification Directive. This point, however, may speak more about the inadequate and controversial nature of Article 8 of the Directive than it does about any shortcoming in the “Hathaway/New Zealand rule”. The Directive sets a minimum standard only and explicitly acknowledges that the primary instrument binding on States is the Refugee Convention. To the non-European reader, the relevance of a regional Directive to the interpretation of an international human rights treaty is problematical at best. As noted by Storey at op cit 7:
The Directive could scarcely do otherwise than leave the Refugee Convention intact, since the individual obligations of the 26 EU Member States under the Refugee Convention remain and have not been revoked. The Directive does not contain any derogations from or renunciations of this Convention or its 1967 Protocol. In any event, international treaty obligations - in this case under the Refugee Convention - cannot be overridden by regional treaties [footnotes omitted].
Rather than approaching the Refugee Convention through the prism of the Qualification Directive we prefer direct application of the Convention in accordance with accepted standards of treaty interpretation as set out in the Vienna Convention on the Law of Treaties, 1969.
Furthermore it is not (and cannot) be suggested that the Qualification Directive is evidence of state practice. As the recent study by the UNHCR of the Qualification Directive (Asylum in the European Union: A Study of the Implementation of the Qualification Directive) establishes, in relation to key provisions of the Directive, “state practice” within the EU is widely divergent. At p 17 the authors of the report are highly critical of Article 8 and recommend that it should be “considered with caution”. The divergent interpretations have also led the authors to recommend at p 65 that decision-makers receive “further guidance as to the benchmarks for the ‘reasonableness analysis’”. In addition the UNHCR has subsequently recommended amendment of some important provisions of the Qualification Directive: Remarks by Ms Erika Feller, Assistant UN High Commissioner for Refugees, Public Hearings On The Future Common European Asylum System, 7 November 2007, Brussels (2008) IJRL 216, 219. In relation to the internal protection alternative in particular, the UNHCR has recommended the deletion of Article 8(3) which permits a determination that an applicant is not in need of international protection notwithstanding technical obstacles to return to the country of origin as well as the amendment of Article 8(1) to require that any proposed area of internal protection be practically, safely and legally accessible to the applicant: UNHCR, Building a Europe of Asylum: UNHCR’s Recommendations to France for its European Union Presidency (July-December 2008) (9 June 2008).
2. That the “rule” is not supported by “such uniformity of international practice” as would establish a rule of customary international law. As to this, it has never been said that the “rule” is a “rule of customary international law”. None of the approaches to the internal flight alternative/relocation/internal protection alternative can claim to be a rule of customary international law. It is precisely because international practice is so divergent that there is an urgent need for a principled approach which will dependably generate consistent rights-regarding protection determinations.
3. That adoption of the rule would give the Convention an effect which is not only unintended, but also anomalous in its consequences:
[19] It would be strange if the accident of persecution were to entitle him to escape, not only from that persecution, but from the deprivation to which his home country is subject. It would, of course, be different if the lack of respect for human rights posed threats to his life or exposed him to the risk of inhuman or degrading treatment or punishment.
As to this, the obligation of non-refoulement applies where there is a well-founded risk of the claimant being persecuted for a Convention reason and the balance of the rights regime in Articles 2 to 33 is also engaged. The Convention does not regard as anomalous that “the accident of persecution” entitles the refugee to escape, not only from that risk of being persecuted, but also from the deprivation to which his home country is subject. This was explicitly conceded by Lord Brown in AH (Sudan) at [41].
Januzi and the re-instatement of human rights
[150] Ironically, having rejected the “Hathaway/New Zealand rule” [that meaningful state protection requires the provision of basic norms of civil, political and socio-economic rights] and having articulated an alternative test based on “reasonableness and undue harshness”, Lord Bingham found it necessary to concede at para [20] of Januzi that decision-makers “should have some guidance” in applying that test. Decision-makers were referred to the “valuable guidance” found in the UNHCR Guidelines of 23 July 2003 which state that the reasonableness analysis is approached by asking “Can the claimant, in the context of the country concerned, lead a relatively normal life without facing undue hardship”? If not, it would not be reasonable to expect the person to move there. In development of this analysis the Guidelines address at para 28 respect for human rights and at paras 29 and 30 economic survival. In essence, the Guidelines state, in effect, that where respect for basic human rights standards is problematic, the proposed area cannot be considered a reasonable alternative. Further, the socio-economic conditions in the proposed area will be relevant as part of the reasonableness analysis.
[151] There is a disconnect between criticising the “Hathaway/New Zealand rule” for requiring account to be taken of basic norms of civil, political and socio-economic human rights and commending an alternative inquiry which arrives at much the same result. It would seem that the “Hathaway/New Zealand rule” is re-instated, if not in name.
[152] Given Lord Bingham’s endorsement of “guidance” which unambiguously accepts the need for a minimum level of human rights to be respected in the proposed area of relocation, it is difficult to understand their Lordship’s strong objection to the “Hathaway/New Zealand rule”. It is equally difficult to understand the objection to informing the internal protection analysis by reference to the Convention’s own minimum standard of rights.
[153] The difficulties inherent in the Januzi decision do not end there. Although all their Lordships agreed with the opinion delivered by Lord Bingham, there appeared to be little unanimity as to the practical application of the reasonableness and undue harshness test. Whereas Lord Bingham began his opinion with the proposition that no account can be taken of any disparity between the civil, political and socio-economic rights which a claimant would enjoy under the leading human rights conventions and those which he or she would enjoy at the place of relocation, he eventually arrived at the position which concedes that the reasonableness analysis can be approached by asking whether the claimant is able to lead a “relatively normal life”; an analysis in which respect for human rights, including the economic survival of the individual, is relevant.
[154] Lord Hope at [59] (Lords Carswell and Mance agreeing), on the other hand, was of the opinion that the almost total absence of civil, political and socio-economic rights was not in itself ground for holding that it would be unduly harsh for the refugee claimant to move to a place of “relocation”. In his opinion only “risk to [the] most basic human rights would rule out the viability of a relocation alternative”. As to this, it could be said, with the greatest of respect, that it is difficult to see the distinction between “risk to [the] most basic human rights” which would rule out relocation and the “almost total absence of civil, political and socio-economic rights” which would not.
[155] With Lord Carswell at [67] stressing “the rigorous nature of the test for unreasonableness or undue harshness”, it seems that the position actually arrived at by several of their Lordships in Januzi is at a substantial distance from the UNHCR Guidelines.
[156] So when the cases of Hamid, Gaafar and Mohammed were reheard by the Asylum and Immigration Tribunal as directed by the Januzi decision, that Tribunal held that even though the refugee claimants would face major violations of their human rights including forced relocations, their situation in Khartoum would not be unduly harsh because the conditions faced by returning Darfuris, however appalling, would be no worse than those faced by other Sudanese internally displaced persons.
[157] On appeal the Court of Appeal held that, on its reading of the Tribunal’s reasons, the Tribunal had misdirected itself in law by setting the test of undue harshness too high, as being conditions so serious as would meet the threshold of ill-treatment under Article 3 of the European Convention for the Protection of Human Rights and Fundamental Freedoms (“ECHR”) (torture or inhuman or degrading treatment or punishment).
[158] The appeal by the Secretary of State to the House of Lords was successful, the House of Lords taking the view that the Court of Appeal had misread the decision of the Tribunal: AH (Sudan) v Secretary of State for the Home Department [2007] 3 WLR 832 (HL) (Lords Bingham, Hoffmann and Hope, Baroness Hale and Lord Brown). The House of Lords did, however, agree that the test of reasonableness, though stringent, was not to be equated with a real risk that the claimant would be subject to such inhuman or degrading treatment as would infringe the rights under Article 3 of the ECHR.
[159] The opinions given in AH (Sudan) are instructive for their commentary on Januzi.
[160] Lord Bingham (with whom Lords Hoffmann and Hope, Baroness Hale and Lord Brown agreed) at [3] began by stating that Januzi was authority for the proposition that in judging reasonableness and undue harshness, no account should be taken of any disparity between the civil, political and socio-economic rights which a claimant would enjoy under the leading international human rights conventions and covenants and those which he or she would enjoy at the place of relocation. Excepted from this rule were breaches of fundamental rights such as those protected by Articles 2 (right to life) and 3 (prohibition of torture) of the ECHR. He then identified para [21] of his opinion in Januzi as containing the correct approach to the internal “relocation” analysis, an analysis depending on “a fair assessment of the facts”:
The decision-maker, taking account of all relevant circumstances pertaining to the claimant and his country of origin, must decide whether it is reasonable to expect the claimant to relocate or whether it would be unduly harsh to expect him to do so ... The decision maker must do his best to decide, on such material as is available, where on the spectrum the particular case falls ... All must depend on a fair assessment of the relevant facts.
After observing in AH (Sudan) at [5] that it was not easy to see how the rule could be more “simply or clearly expressed” Lord Bingham acknowledged that the difficulty lay “in applying the test, not in expressing it”.
[161] Not explained was how a fair assessment of the facts could be made without taking into account basic norms of international human rights beyond Articles 2 and 3 of the ECHR. It is significant that Lord Bingham, having acknowledged the difficulty in applying a test formulated in terms of “reasonableness”, “undue harshness” and “a fair assessment of the relevant facts”, eventually conceded that “very little” could be excluded from consideration, other than the standard of rights protection available in the country of asylum. See para [13]:
As already indicated (para 5 above) the test propounded by the House in Januzi was one of great generality, excluding from consideration very little other than the standard of rights protection which an applicant would enjoy in the country where refuge is sought.
[162] In our respectful view, once it is accepted that little is excluded from the “relocation” exercise other than the standard of rights protection the individual would enjoy in the State party where refuge is sought, it is difficult to understand why their Lordships in Januzi and AH (Sudan) went to such lengths to reject the relevance of human rights to the inquiry and in particular the modest set of standards suggested by the Refugee Convention itself.
Subjectivity revisited - “standards prevailing generally in the country of origin” and “a relatively normal life”
[163] The subjectivity inherent in the “unduly harsh” test is underlined by the notion of a refugee claimant being able, in the proposed place of relocation, to live “a relatively normal life” in the context of standards prevailing generally in the country of origin (Januzi at para [20] per Lord Bingham and para [47] per Lord Hope):
If the claimant can live a relatively normal life there judged by the standards that prevail in his country of nationality generally, and if he can reach the less hostile part without undue hardship or undue difficulty, it will not be unreasonable to expect him to move there.
[164] In the later AH (Sudan), Lord Brown (with whom Lord Hope agreed) at para [42] explained that the decision-maker is to determine what suffering the refugee claimant is able to bear. Only proof that his or her life would be “quite simply intolerable” would lead to recognition of refugee status:
If a significant minority suffer equivalent hardship to that likely to be suffered by a claimant on relocation and if the claimant is as well able to bear it as most, it may well be appropriate to refuse him international protection ...
... only proof that their lives on return would be quite simply intolerable compared even to the problems and deprivations of so many of their fellow countrymen would entitle them to refugee status. [Emphasis added]
[165] Not explained is how a decision-maker is to realistically assess whether the claimant can live a relatively normal life judged by the standards prevailing generally in the country of nationality; whether the claimant is “able to bear” his or her hardship “as most” and the point at which life “would be quite simply intolerable”.
[166] The dangers inherent in a test couched in such subjective terms are illustrated by AA (Uganda) v Secretary of State for the Home Department [2008] EWCA Civ 579 (22 May 2008) (Buxton, Carnwath & Lloyd LJJ). While this was a claim to protection under Articles 3 and 8 of the ECHR, the case was disposed of by the English Court of Appeal on the “relocation” principles set out in Januzi and AH (Sudan). Ms AA was a citizen of Uganda and it was accepted that she would be at serious risk of suffering serious harm were she to be returned to northern Uganda. The case turned on whether she could relocate to Kampala where she was without accommodation, had almost no chance of obtaining employment in the formal sector and her only hope was to find work as a prostitute in the slums. The Asylum and Immigration Tribunal had determined that the situation facing Ms AA was the same as that of many other young women living in the slums of Kampala and, quoting from Lord Hope in Januzi (whether the claimant could live a relatively normal life judged by the standards that prevail in the country of nationality generally), found that it would not be unduly harsh for Ms AA to relocate to Kampala.
[167] The decision of the Tribunal was set aside. Buxton LJ (the other members of the Court agreeing) explained AH (Sudan) on the basis that in that case the conditions in the place of relocation involved poverty, disease and the living of a life that was structured differently from that which the appellants had come from in Darfur. Nevertheless it had been open to the Tribunal in that case to hold that exposure to those conditions, shared by many of the refugees’ fellow-countrymen, did not amount to undue harshness. In the case of Ms AA, however, she was faced not merely with poverty and lack of any sort of accommodation, but with being driven into prostitution. At para [17] he continued:
Even if that is the likely fate of many of her fellow countrywomen, I cannot think that either the AIT or the House of Lords that decided AH (Sudan) would have felt able to regard enforced prostitution as coming within the category of normal country conditions that the refugee must be expected to put up with. Quite simply, there must be some conditions in the place of relocation that are unacceptable to the extent that it would be unduly harsh to return the applicant to them even if the conditions are widespread in the place of relocation.
[168] Carnwath LJ at para [41], explaining that it was unnecessary to explore the application of Article 3 of the ECHR, stated:
Common law principles are sufficiently robust for us to hold that in all the circumstances of this case a decision to return this claimant to Kampala would be “so outrageous in its defiance ... of accepted moral standards” (see CCSU v Minister of Civil Service [1985] AC 374, 410G) that it could not lawfully be upheld.
[169] The English Court of Appeal was unquestionably right in setting aside the decision of the Tribunal but the outcome should not have depended on the distinction between living in poverty and disease on the one hand and living in poverty, disease and enforced prostitution on the other. In our respectful view, rather than the issue turning on what is (or is not) “acceptable” or on whether the circumstances are so outrageous as to defy accepted moral standards, the analysis should be based on human rights standards, including the right to security of person; freedom from deprivation, including access to food and shelter; protection of property; freedom of thought, conscience and religion; access to basic education; documentation of identity and status; access to the courts and respect for family unity. All of these rights are expressly or impliedly recognised by the Refugee Convention and would allow a more objective and consistent determination of the difficult decisions to be made in the internal protection alternative context.
[170] The hazardous nature of an approach based on “reasonableness” and “undue harshness” is highlighted also by Baroness Hale in AH (Sudan) at paras [27] to [29] where she points out that Januzi had been understood by the Asylum and Immigration Tribunal as requiring comparison with the lives of “the poorest of the poor” internally displaced victims of the civil war:
[27] ... We know that the standard of comparison is not the lives which the returning claimants are living here: that is what Januzi was all about. We know that the lives they led before the persecution are a relevant factor but not, as the Court of Appeal thought, the starting point. We know that the lives they will face on return have to be considered in the context of “standards prevailing generally in the country of nationality”: Lord Bingham in Januzi [2006] 2 AC 426, para 20. If people can return to live a life which is normal in that context, and free from the well-founded fear of persecution, they cannot take advantage of past persecution to achieve a better life in the country to which they have fled: see Lord Bingham in para 5 of his opinion. But this does not mean that the holistic consideration of all the relevant factors, looked at cumulatively, can be replaced by a consideration of whether their circumstances will be worse than the circumstances of anyone else in that country. [Emphasis in original]
[28] Yet the tribunal concluded that because the conditions faced by returning Darfuris, however appalling, would be no worse than those faced by other Sudanese IDPs it would not be not “unduly harsh” to expect them to return. The standard of comparison was, not with their lives in Darfur before their persecution, not with the general run of ordinary lives in Sudan, not even with the lives of poor people in Sudan, but with the lives of the poorest of the poor internally displaced victims of the civil war in the south, living in camps or squatter slums, and “subject from time to time to relocations, sometimes involving force and human rights violations” ... They too had been subsistence farmers, ill-equipped to survive in the city slums ...; they too had suffered the psychological horrors of civil war ..., if not of government-backed genocide; the Darfuris would be no worse off, unless particular individuals attracted the adverse interest of the authorities: .... With respect this is not the individualised, holistic assessment which the question requires.
[171] As Hathaway and Foster at op cit 385 point out, the “reasonableness” test in practice allows decision-makers to assess the asylum-seeker’s alternatives in light of their own view of what constitutes “reasonable” behaviour. They illustrate the point by citing from Frelick, “Down the Rabbit Hole: The Strange Logic of Internal Flight Alternative”, World Refugee Survey (US Committee for Refugees, Washington DC, 1999) 22, p 23:
Reasonableness, as Alice no doubt would observe, depends on which side of the looking glass one is standing. Viewed from the host country perspective, the risks and dangers to asylum seekers back in their far away countries may appear less threatening than they do from the perspective of persons who have directly experienced those conditions up close and who fear being sent back through the looking glass to experience them again.
[172] The “reasonableness” test has also been criticised by Kirby J in SZFDV at [44] and [45]. Although he was not there addressing directly the contrast between the “reasonableness” and “protection” approaches, his criticism of a “relocation” test framed in terms of reasonableness is nonetheless a telling one. He points out that to frame the test in terms of “reasonableness” or “undue harshness” opens “an all too easy exit” from the assessment required by a test framed in protection terms:
[45] If the present appeal is dismissed, that “easy exit” will be invoked in even more cases in the future. This will be especially so where a country is large and diverse and where the refugee claim adjudicator or the tribunal simply postulates their own view as to the reasonableness of an applicant’s internal relocation. That postulate should never be allowed to undermine the important rights expressed in the Refugee’s Convention....
[173] Illustration of this point is provided by the UNHCR, Asylum in the European Union: A Study of the Implementation of the Qualification Directive (November 2007) at p 60 where it is noted that in Germany many Chechen asylum-seekers are denied refugee status on the grounds that they could live elsewhere in the Russian Federation, the decision-makers being of the view that the lack of a legal right to reside or to work in the proposed location does not negate the finding of an internal relocation alternative (pp 60-61):
Even though it was acknowledged that they might not attain a basic level of subsistence in the alternative region in the Russian Federation, as the economic situation was worse in Chechnya, an internal protection alternative was affirmed.
As a consequence, Chechen applicants are expected to settle in other areas of the Russian Federation without official registration, and dependent on the informal labour market to earn their living.
[174] In a protection-focused analysis such outcomes would not be possible. But they are possible in a reasonableness analysis which is not anchored in the language, object and purpose of the Refugee Convention. It involves an unfocused and open-ended inquiry and is prone to arbitrariness and subjective interpretation by the decision-maker.
[175] In contrast, locating the inquiry in the protection element of the definition means that putative refugee status is lost only if the individual can access in his or her own country of origin the same level of protection that he or she would be entitled to under the Refugee Convention. See Refugee Appeal No. 71684/99 at [57] to [61]. As the Michigan Guidelines on the Internal Protection Alternative explain at paras [20] to [22]:
20. The denial of refugee status is predicated not simply on the absence of a risk of persecution in some part of the state of origin, but on a finding that the asylum-seeker can access internal protection there. This understanding follows from the prima facie need for international refugee protection of all asylum-seekers whose cases are subjected to internal protection analysis. If recognition of refugee status is to be denied to such persons on the grounds that the protection to which they are presumptively entitled can in fact be accessed within their own state, then the sufficiency of that internal protection is logically measured by reference to the scope of the protection which refugee law guarantees.
21. Good reasons may be advanced to refer to a range of widely recognized international human rights in defining the irreducible core content of affirmative protection in the proposed site of internal protection. In particular, one might rely on the reference in the Refugee Convention's Preamble to the importance of '... the principle that human beings shall enjoy fundamental rights and freedoms without discrimination'. Yet the Refugee Convention itself does not establish a duty on state parties to guarantee all such rights and freedoms to refugees. Instead, Arts. 2-33 establish an endogenous definition of the rights and freedoms viewed as requisite to '... revise and consolidate previous international agreements relating to the status of refugees and to extend the scope of and the protection accorded by such instruments ... (emphasis added)'. These rights are for the most part framed in relative terms, effectively mandating a general duty of non-discrimination as between refugees and others.
22. At a minimum, therefore, conditions in the proposed site of internal protection ought to satisfy the affirmative, yet relative, standards set by this textually explicit definition of the content of protection. The relevant measure is the treatment of other persons in the proposed site of internal protection, not in the putative asylum country. Thus, internal protection requires not only protection against the risk of persecution, but also the assimilation of the asylum-seeker with others in the site of internal protection for purposes of access to, for example, employment, public welfare, and education.
[176] As these paragraphs make clear, the refugee-specific statement of rights found in the Refugee Convention is used to measure the content of protection in the proposed site of safety in the country of origin. The relevant measure is the treatment of other persons in the proposed site of internal protection, not in the putative asylum country. As Hathaway and Foster at op cit 409 point out, it is important to understand that this approach does not suggest a literal application of Articles 2 to 33 in considering internal protection, but rather that decision-makers seek inspiration from the kind of interests protected by these articles as a way of defining an endogenous notion of affirmative protection in the refugee context. While in some ways falling somewhat short of the standard of “protection” that would follow from assessment by reference to all key human rights or even to the “Guiding Principles on Internal Displacement”, the non-discrimination approach embodied in the Refugee Convention nonetheless provides a legally solid and contextualised assurance of durable protection.
Januzi and AH (Sudan) - conclusions
[177] Even were the Authority free to depart from Butler (which it is not), it would not follow Januzi and AH (Sudan). Our two broad reasons are:
(a) The failure of these decisions to recognise that, as a matter of treaty interpretation, the textual location of the internal protection alternative analysis is not in the well-founded element of the refugee definition, but in the protection element.
(b) Their failure to recognise that the inquiry must be seen as a protection inquiry, not one of “reasonableness” or “undue harshness” or of what is “intolerable” or what the claimant “is able to bear”.
[178] In these circumstances the Authority affirms the “Hathaway/New Zealand rule”, namely that once a refugee claimant has established a well-founded fear of being persecuted for a Convention reason, recognition of that person as a Convention refugee can only be withheld if that person can genuinely access in his or her home country domestic protection which is meaningful. Such protection is to be understood as requiring:
(a) That the proposed internal protection alternative is accessible to the individual. This requires that the access be practical, safe and legal.
(b) That in the proposed site of internal protection there is no risk of being persecuted for a Convention reason.
(c) That in the proposed site of internal protection there are no new risks of being persecuted or of being exposed to other forms of serious harm or of refoulement.
(d) That in the proposed site of internal protection basic norms of civil, political and socio-economic rights will be provided by the State. In this inquiry reference is to be made to the human rights standards suggested by the Refugee Convention itself.
[179] It is now possible to turn to the question whether recognition of the appellant’s status as a Convention refugee can properly be withheld on the basis that she has an internal protection alternative.
[180] As to the first step of the inquiry (access), it will be assumed, without deciding, that the appellant will have practical, safe and legal freedom of movement within Turkey. This assumption is made largely because, as will be seen, the substantive impediments to her accessing meaningful state protection lie in the second, third and fourth steps to the inquiry.
[181] As to the second step of the inquiry (the existence of an “antidote” to the risk of being persecuted), the appellant presents as a forty-year-old woman who faces the prospect of return to Turkey as a single divorced woman without a male protector. Her two sons presently aged 21 and 20 will remain in New Zealand where they have been recognised as Convention refugees. In Turkey the appellant will be at risk of being killed either by her husband’s family or by her own family, or by both. It will not be possible for her to find assistance within the Kurdish community as it is largely from within that community that the threat to her safety comes. As a single divorced woman without a male protector she will have no opportunity to “blend” into such Kurdish community as she might find herself in and there is a real risk that the fact of her return will eventually be reported to her family or to that of her husband.
[182] Because of the risk posed by members of the Kurdish community she would have to find a place of internal protection outside the Kurdish areas of Turkey, that is away from the east and southeast. In practical terms this leaves only the major population centres. However, decades of conflict between the Turkish security forces and Kurdish insurgents have caused substantial internal displacement of Kurds to these same population centres. Istanbul in particular is not a viable option given that this city is the home of the brother who has threatened to kill her. This leaves the appellant with few practical options. There is no realistic prospect of her finding a place in a shelter for female victims of domestic violence as there are few such shelters in relation to the number of women in need of assistance. Even were such shelter to be available she would not receive effective protection by the state from the threat posed by her relatives and their community. The country information establishes that at every level of the criminal justice system the authorities fail to respond promptly or vigorously to women’s complaints of family violence, including the violent deaths of women. See the report from Amnesty International referred to earlier, Turkey: Women Confronting Family Violence at 18 and 19. In addition it is reported that despite government efforts, honour-related crimes show little sign of abating: “A Dishonourable Practice” The Economist (April 14, 2007) 57.
[183] In summary, there is real difficulty in identifying a geographical location in Turkey where the appellant would be free of a real chance risk of being killed or seriously harmed for her alleged transgression of the “honour” code. The exacting internal protection alternative test as applied in New Zealand precludes a decision-maker from assuming, without adequate evidentiary basis, that somehow the putative refugee will muddle her way to a geographical location in which she is free from the risk of being persecuted for a Convention reason. In the circumstances the Authority concludes that it is not possible to identify a site of internal protection which will provide an antidote to the well-founded fear of being persecuted, that is a place where the appellant will not be at risk of being persecuted for a Convention reason.
[184] As to the third step in the inquiry (exposure to other harm), it is not possible to say that there is a place of protection in Turkey where there will be no new risk of the appellant being persecuted or of being exposed to other forms of serious harm. The appellant will face multiple layers of discrimination. First, because of her sex and gender, second, because she is a Kurd, third, because of her age and finally, because of her lack of education and marketable work skills. These layers of discrimination will severely handicap her finding shelter, employment and food. Having only primary education and never having been in paid employment, she will have almost no ability to support herself in Turkey either in the formal or in the informal employment sector; especially given that participation by women in the Turkish workforce is among the lowest in the OECD countries and that there is prevailing discrimination in the labour market: Commission of the European Communities, Turkey 2006 Progress Report at pp 17-18 and Turkey 2007 Progress Report at p 18.
[185] In the circumstances there is a real risk that the appellant might feel driven to seek the assistance of her family in the hope that on seeing her predicament they will withdraw the threats to kill her and offer assistance. On the facts known, this would be an unrealistic expectation but if the appellant is faced with a desperate situation, she might elect to accept the risk nonetheless. This is precisely the situation addressed by step 3 of the internal protection alternative inquiry, namely the exposure of the refugee to circumstances which create the risk of indirect refoulement back to the conditions which gave rise to the putative refugee status in the first place.
[186] Given these findings it is largely academic to engage with the fourth step of the inquiry ie whether there will be a minimum level of civil, political and socio-economic rights. But it follows from our earlier findings that the appellant will not enjoy a range of rights recognised by the Refugee Convention, including protection of physical security and access to the necessities of life, particularly food, housing and work. In Turkey, social security benefits are not universal; one needs to be legally employed (or self-employed) or be the dependent of a legally employed (or self-employed) person in order to be covered under a social security programme: Internal Displacement Monitoring Centre, Turkey: Progress on national IDP policy paves way for further reforms - A profile of the internal displacement situation (26 July 2007) at p 118. Exclusion from social security leads, in turn, to exclusion from most forms of heath care: Ibid 97.
[187] It follows that there is no internal protection alternative for the appellant and there are no proper grounds on which recognition of her status as a Convention refugee can be withheld.
[188] Our conclusion that there is no internal protection alternative in Turkey for a woman facing death by honour killing is reinforced by two recent decisions given by administrative courts in Germany and to which the Authority has been given access by the UNHCR office in Nuremberg and the Protection Policy and Legal Advice Section, Division of International Protection Services, UNHCR, Geneva. The decisions in question are of the Stuttgart Administrative Court, Decision of 29 January 2007, A 4K 1877/06 (Turkey) and of the Darmstadt Administrative Court, Decision of 17 October 2007, 8 E 1047/06.A(1) (Women from Turkey of Kurdish Ethnicity). We express our appreciation and gratitude to the UNHCR for providing translated extracts and for facilitating access to decisions which would not otherwise have been available to the Authority. While the jurisprudential path taken in those cases is different from that followed by the Authority, the conclusion on the facts that women at risk of being killed in an honour crime have no internal protection alternative in Turkey is in accord with our own conclusion in relation to the particular circumstances of the present appellant.
[189] For the reasons given we find that the appellant has a well-founded fear of being persecuted in Turkey for reasons of political opinion and membership of a particular social group. There is no internal protection alternative. The appellant is recognised as a refugee. The appeal is allowed.
“Rodger Haines”
.....................................
[Rodger Haines QC]
Chairperson