Refugee Status Appeals Authority
REFUGEE APPEAL NO. 71427/99
AT AUCKLANDBefore: R P G Haines QC (Chairperson)
L Tremewan (Member)
Counsel for the Appellant: Margaret Robins
Date of Hearing: 14 July 1999
Date of Decision: 16 August 2000
IRAN: INEQUALITY OF MARRIAGE, DIVORCE AND CUSTODY RIGHTS
DOMESTIC VIOLENCE IN IRAN
THE APPELLANT’S CASE
ASSESSMENT OF THE APPELLANT’S CREDIBILITY
THE PERSECUTION ISSUE
Understanding the Meaning of Persecution
Discrimination as Persecution
Agents of Persecution
The Standard of State Protection
Is there a Presumption of State Protection?
Whether Appellant faces Persecution
The Serious Harm Issue
The Failure of State Protection Issue
Conclusion on the Persecution Issue
THE CONVENTION GROUNDS
Race and Nationality
Religion and Political Opinion
PARTICULAR SOCIAL GROUP
Whether the Appellant is a Member of a Particular Social Group
THE NEXUS ISSUE
Nexus: The First Husband
Nexus: The State
 This is an appeal against the decision of the Refugee Status Branch of the New Zealand Immigration Service declining the grant of refugee status to the appellant, a citizen of the Islamic Republic of Iran.
 The appellant’s case has raised difficult jurisprudential issues and the delay in delivering this decision is regretted. The inquisitorial responsibilities of the Authority have, however, required the Authority to engage in extensive research. It is largely due to those researches that it has been possible to arrive at a conclusion on both the facts and on the law that the appellant is a refugee within the meaning of Article 1A(2) of the Refugee Convention.
IRAN: INEQUALITY OF MARRIAGE, DIVORCE AND CUSTODY RIGHTS
 This case has its origin in an ill-fated marriage (soon followed by divorce) and a tragic custody battle. An understanding of the appellant’s case requires a brief overview of the institutionalised and state-sanctioned discrimination against women in the Iranian family context. It is a topic the Authority addressed in Refugee Appeal No. 2039/93 Re MN (12 February 1996) at 19-31. The following passage from Ann Elizabeth Mayer, Islam and Human Rights: Tradition and Politics (3rd ed, 1999) 103 provides a general introduction:“It is assumed that all women will marry, so the primary determinant of an adult women’s life will be her relationship with her husband. In this relationship, she is required to submit to her husband’s authority and follow his wishes. It is expected that her life will be passed at home fulfilling domestic duties. There is no concern for protecting women’s rights to develop as individual persons with distinct identities and abilities, to become educated in ways that fit their specific talents and interests or that enable them to become productive members of society, or ensure that they play a part in the social, economic, or political institutions that shape their destinies. Women are seen not as actors but as passive, dependent beings - all of whom are basically fungible, not diverging in personality and capacity as males do.” On the specific issues of marriage, divorce and custody, the following points have been taken, unless otherwise indicated, from the study by Sima Pakzad, “The Legal Status of Women in the Family in Iran” in Mahnaz Afkhami & Erika Friedl (eds), In The Eye of the Storm: Women in Post-Revolutionary Iran (Tauris, 1994) 169-179:
(a) A woman is unable to retain and propagate her family name once she becomes married, op cit 169. The mother cannot, even with her husband’s consent, confer her maiden name on her child. The law is both a reason for and a reflection of the fact that in most Iranian families a male child is held in higher regard and that the ultimate wish of most fathers is to have a son who will be the bearer and guardian of their name.
(b) The minimum age of majority is fifteen lunar years for boys and nine lunar years for girls, op cit 169-170. The latter age was eighteen prior to 1979 but was reduced after Khomeini’s accession to power. The marriage of the girl, even after she has reached the age of majority, depends on the consent of her father or her paternal grandfather.
(c) A daughter’s inheritance from her father or mother is half that of a son’s share, op cit 170.
(d) Citizenship is granted on the basis of one’s paternal status. Those born in and outside Iran and whose fathers are Iranian are considered Iranian citizens. But if only the mother of the child is Iranian, the child will be granted Iranian citizenship provided that the child has been born in Iran and has resided at least one more year in Iran immediately after reaching the full age of eighteen. If a man applies for Iranian citizenship and his application is approved his minor children also become Iranian. However, if a women becomes a naturalized Iranian citizen, her minor children are not considered, ipso facto, Iranians, op cit 171.
(e) Marriage, according to the current Iranian law, consists of two kinds - permanent and temporary. In permanent marriage, as the name indicates, no duration is specified in the marriage contract. Temporary marriage, also known as sigheh, on the other hand, can last only for a specific period of time. In permanent marriage, a wife enjoys a higher degree of security and respect within the family. In temporary marriage, on the other hand, matrimonial relations are considered terminated and the wife must leave the husband’s residence as soon as the specified period is over or if the husband waives his right to the remaining portion of the said period. In addition, while in permanent marriage the husband is legally responsible for supporting his wife, in temporary marriage the wife is not entitled to such support, op cit 172.
(f) Following the act of marriage, the wife must live in the dwelling that the husband designates. By leaving her husband’s residence in order to live in another dwelling, including her father’s, the wife loses the right to financial support, op cit 172-173.
(g) Marriage of a Muslim woman with a non-Muslim man is prohibited, that is, a Muslim woman may only marry a Muslim man. There is no such limitation on the marriage of the Muslim man, op cit 173.
(h) Iranian nationality is granted to any woman of foreign nationality who marries an Iranian man. However, if an Iranian woman marries a man of foreign nationality, not only is her husband not accorded Iranian citizenship but she herself may be, depending on the nationality laws of her husband’s country of nationality, be compelled to acquire the same nationality as that of her husband, in which case she will lose her Iranian citizenship, op cit 174.
(i) A married woman may, without requiring her husband’s permission, be gainfully employed. However, if the nature of her occupation is not compatible with the family’s interests or dignity, her husband may prevent his wife from engaging in such an occupation, op cit 174.
(j) An Iranian man may take several wives and there are no limits to the number of temporary wives that he is permitted to have. While Islamic law permits polygamy only if the husband is able to be equally fair to all of his wives, under the Civil Code the husband is the sole judge of the issue, op cit 175 .
(k) To leave Iran, a married woman needs her husband’s written permission, op cit 175.
(l) The husband may divorce his wife whenever he wishes, whereas a woman may initiate divorce only under certain limited conditions. Generally divorce is possible if the continuation of the marriage causes undue hardship. In general this is interpreted as including any unsavoury development or occurrence such as the husband’s drug addiction, his association with unsavoury characters, his contraction of a serious contagious disease or his imprisonment, op cit 178.
(m) Inheritance in the context of permanent marriage is unequal. The husband’s share is one-quarter of the estate if the wife is survived by any children or grandchildren. If she is not so survived the husband’s share is one-half of the estate. But if the wife should survive the husband, her respective shares are one-eighth and one-quarter respectively. In the case of a polygamist husband, the wives share equally the one-eighth and one-quarter shares as the case may be. Furthermore, while the husband’s share of the inheritance covers all of the wife’s property, the wife’s share of the estate is only from movable property, buildings and trees, op cit 178-179.
(n) The concept of “custody” has a double meaning in Shi’i shariat. The first meaning refers to legal guardianship (velayat). This is a right which naturally and automatically belongs to the father and paternal grandfather in his absence. The second meaning refers to fostering (hezanat), that is caring for offspring for a fixed period of time without possessing legal guardianship. The former meaning refers to the right of the man, while the latter is considered as a natural, but not automatic, right of the mother and extends in the case of a female child until the age of seven and the case of a male child up to the age of two. The right to foster can be withdrawn from the mother if her moral suitability is in doubt: Parvin Paidar, Women and the Political Process in Twentieth-Century Iran (Cambridge University Press, 1995) 294.
 The text by Parvin Paidar, Women and the Political Process in Twentieth-Century Iran (Cambridge University Press, 1995), and in particular Chapter 8 entitled “The Islamic Construction of Family” 267-302, is particularly instructive as to the socio-political context in which the gender policies of the Islamic Republic have been developed and implemented. Also useful is the brief by the Research Directorate, Documentation, Information and Research Branch, Immigration and Refugee Board, Ottawa, Canada, Human Rights Brief: Women in the Islamic Republic of Iran (June 1994) para 2.6.
 The depth of gender-based discrimination in the Iranian legal system, and indeed in the society at large, is underlined by the fact that gender discrimination is also the central feature of the Iranian penal code. A convenient summary entitled “The Islamic Penal Code of the Islamic Republic of Iran: Excerpts Relating to Women” is reproduced in Mahnaz Afkhami & Erika Friedl (eds), In The Eye of the Storm: Women in Post-Revolutionary Iran (Tauris, 1994), Appendix II, 180-187. We mention only a few examples. First, the value of blood money, which is based on how much a person would have earned in a lifetime, is twice as much in the case of a murdered man as in the case of a woman. As some have pointed out, the life of a woman who is murdered is worth only half that of a man’s. The penal code also stipulates that the number of witnesses required to prove a crime is higher if the witnesses are female. In the case of adultery the testimony of women alone or in conjunction with the testimony of only one man cannot prove adultery but rather, constitutes a false accusation which itself is a punishable act. The testimony of four men or three men and two women is required to prove adultery. First degree murder must be proven by the testimony of two men. Evidence for second degree murder or manslaughter requires the testimony of two men, or one man and two women, or the testimony of one man and the sworn testimony of the accuser.
 The significance of this state-sanctioned inequality of women within the context of the Islamic theocratic regime in Iran is discussed in Refugee Appeal No. 2039/93 Re MN (12 February 1996). We will return to this decision later.
DOMESTIC VIOLENCE IN IRAN
 Domestic violence is regrettably endemic in most societies and Iran is no exception. We do not intend to recite the evidence at length. It is sufficient to note that in Parvin Paidar, Women and the Political Process in Twentieth-Century Iran (Cambridge University Press, 1995) 352 it is recorded that according to research conducted in Tehran on divorce and its causes, 75 per cent of women who were involved in divorce disputes said that they had been physically attacked by their husbands. The rate of domestic violence was reported to be as high as 72 per cent. Many incidents of domestic violence ended with a woman’s death or serious injury. In Chapter 8 entitled “The Islamic Construction Of Family”, Paidar explains at op cit 267-302 how the clerical construct of the “Islamic family” has fostered domestic violence in Iran. The conclusion drawn is that Muslim women are failed by the Islamic state’s policies and by the courts of law.
 The state has even taken steps to outlaw public discussion of domestic violence. See Julian Borger, “Women in Iran targeted by new laws” Guardian Weekly, May 24, 1998 p 5. This article reports that in the preceding week the Majlis (the Islamic consultative assembly) had finalised two bills which if passed, would outlaw press coverage of domestic violence, stifle criticism of laws affecting women and segregate medical services. The ban on press coverage of domestic violence was aimed at Iran’s increasingly varied range of newspapers and to stifle growing debate over the application to women of Islamic law. It has not been possible to ascertain whether the legislation has been passed, but the bill did receive a second reading in August 1998: Human Rights Watch World Report 1999: Iran 352, 356. The only other information to hand is the mention in the Department of State, Country Reports on Human Rights Practices for 1998: Iran (April 1999) 1660 at 1671, that the bill providing for the segregation of medical services was initially rejected by the Council of Guardians pending an amendment to assure funding, but the bill was subsequently approved by the Council in a review in November 1998. While the Authority has not been able to ascertain whether the ban on press coverage of domestic violence was approved by the Council of Guardians, the fact that the Majlis even drafted such legislation is revealing of the attitude of the Iranian state to the issue.
 It is plain from the country information that the attitude to domestic violence by the Iranian state is one of condonation, if not complicity. Usually, a file will only be opened if a woman produces a certificate of serious injury for a second or third time. Furthermore, the Iranian courts (such as they are) know of no such thing as a restraining order against a husband. Nor are there laws to prevent stalking. See generally the Response to Information Request (REFINFO IRN29616.E) Iran: Information on the way women seek to protect themselves from spousal abuse, under what circumstances family, friends, neighbours, and the religious authorities would intervene when a woman is being abused by her husband, under what circumstances the police would intervene, and whether a married woman or her family can obtain a restraining order against an abusive husband (1998/07/00):“The following information on police intervention in cases of domestic abuse was provided to the Research Directorate by a professor of sociology at Concordia University in Montreal, who specializes in women’s issues in Iran. If a woman who is being beaten by her husband calls the police from her home, it is unlikely that they would intervene; however, the woman has the option of going to the police station to lodge a complaint against her husband. If a woman chooses that option, she must produce a medical certificate proving that she has received a serious physical injury (eg a broken bone or knife-wound) at the hands of her husband before the police will open a file on the case. Moreover, the professor is not aware of any instance in which the police have opened a file after the first certified instance of physical injury at the hands of a husband. However, the professor believes that the police will usually open a file if a woman produces a certificate of serious injury for a second or third time ( 26 June 1998).Regarding the issuance of restraining orders by Iranian courts, the professor said that there is no such thing as a restraining order against a husband in a case where the wife is living with the husband. However, if a married woman can prove to a court that her life is in danger from her husband, the court may allow her to move to her father’s home, in which case the husband can be kept away legally (ibid.).This appears to be corroborated by an article on Iran’s Personal Status Law, published in 1996. According to the authors, In the earlier REFINFO IRN26714.E Iran: Update to Responses to Information Requests IRN19097.E dated 19 December 1994 and IRN16039.E dated 27 January 1994 regarding domestic violence and protection available to women victims of spousal abuse (1997/05/00) it is recorded that in an interview with a specialist on women’s issues in Iran at the Department of Sociology at York University, Toronto, it was revealed that the source was not aware of state-run services for women victims of spousal abuse and the source also doubted that the police would take a complaint of domestic violence seriously.A wife may choose to live in a separate residence if she can prove to the court that she has a reasonable fear of physical harm or harm to her reputation. If the court accepts her claim, she is entitled to receive economic support (nafaqa) until the couple reaches an agreement or the marriage ends. However, despite these provisions, it is very difficult for women to convince the court that they are in danger from their husbands. Prior history of abuse is considered evidence of danger only if the battery has caused major injury; this suggests that battery is permitted as long as it does not result in permanent harm or handicap. The judge is left to rule on the severity of a situation, and decisions are highly subjective. (WLUML 1996, 20)”
THE APPELLANT’S CASE
 The appellant is a 32 year old woman who has one child, a son now aged 11. In circumstances about to be described, the appellant was divorced from her first husband in June 1989. She subsequently entered into a temporary marriage with her current partner (presently still in Iran) in April 1997.
 She is the second eldest in a family comprising her mother, father, five daughters and one son. The home environment was described as traditional but relaxed. When the appellant was 13 years old the family moved from Tehran to Karaj.
 Arranged marriages appear to be the norm in the appellant’s family. Four of the five daughters have entered into such marriages and the appellant expects that the same will happen to her remaining sister. In the appellant’s case, she was 19 years of age when her parents arranged a marriage with an apparently religious man called [name deleted] (“the first husband”) who is also a member of the Sepah Pasdaran (Revolutionary Guards). At the time she saw nothing unusual in the fact that she first met her husband-to-be one month before the marriage and knew virtually nothing about him. She trusted in the judgment of her father. The groom came from a conservative Muslim family and was four years her senior. The appellant’s mother was particularly supportive of the marriage because she thought that the religious values held by the first husband and his family would have a positive effect on her daughter.
 After the marriage in November 1987 the appellant and her first husband moved into an apartment near the husband’s family home. From that point the appellant became a prisoner, forbidden from leaving the house in her husband’s absence and was locked inside when he went out. She was expected to adhere to strict “Islamic” rules which included complying with the dress code even indoors and observing protocols relating to the segregation of males and females. Her husband also objected to her visiting her brother on the grounds that the brother was of the male gender. The appellant noted that all the women in her husband’s family were devout Muslims who complied unquestioningly with the directions given by male members of the family. As a result she did not see her family often and they did not realise that she was unhappy.
 Soon after the marriage the appellant began to be beaten regularly. When, several months after the marriage, the appellant discovered that she was pregnant her husband told her that he did not care about her and refused to allow her out of the household for tests and checkups, notwithstanding that the appellant was suffering from severe back pain as a result of her pregnancy. Once the husband learnt that his wife was pregnant, the beatings intensified.
 In the seventh month of her pregnancy the appellant found in her husband’s wallet the photograph of a woman. When she questioned him about this discovery, he told her that he loved the woman concerned and wanted to marry her. When the appellant’s mother-in-law visited a short time later and was told by the appellant of what she had found, the mother-in-law accused her of lying.
 Once the husband’s affair was in the open, he brought the other woman home and required the appellant to treat her as a guest. The fact that the appellant was pregnant did not appear to concern either the husband or the girlfriend. The appellant was locked in a room while the pair had sex.
 When in August 1988 the appellant started to go into labour her husband refused to take her to hospital, saying that he wished she and the baby would die so that he would be free of them. Finally, he sent for his mother who took the appellant to hospital.
 After the birth of the appellant’s son, the mother-in-law gave instructions through relatives who worked at the hospital that the appellant was not to be shown the child and was to be told only that the baby had contracted hepatitis B and had to be kept at the hospital. After two days the appellant was discharged without seeing her child. A week later she was told by her mother-in-law that the boy had died. The appellant did not know at the time that this was a lie.
 A month later the appellant was told by her first husband that there was no reason for her to continue living with him as he did not love her and there was no baby to look after. He sent the appellant back to her parents who, until then, had had no idea that their daughter had been subjected to violence and kept prisoner for almost one year.
 Without the appellant’s knowledge, her husband sold the baby to a couple, a Mr and Mrs [X], who were unaware that the child’s mother was alive and well.
 Over the next year no progress was made in the divorce proceedings. It was only later that the appellant discovered that her first husband had brought his influence to bear on the family court in order to delay matters in the hope that the appellant would relinquish her rights under the marriage settlement. In the end the appellant went to the High Court in Tehran where her husband had no influence. Following the intervention of that court the husband finally appeared in the Karaj family court in June 1989. It was at this hearing, when the divorce was formalised, that the appellant discovered for this first time that her child was alive. Simultaneously, the court ordered that the husband be given custody of the child.
 The first husband proved to be vindictive even after the divorce and instigated Pasdars under his control to embark on a campaign of harassment of his former wife. This resulted in the appellant being repeatedly arrested for allegedly violating the dress code and the code which requires the separation of male and female. By way of example the appellant was on one occasion on her way to a funeral in the company of a male family friend when they were both detained, taken to the Komiteh and flogged 20 times.
 Exhausted by two years of emotional and physical punishment, the appellant became ill and was hospitalised for a short time. After slowly regaining her health she began to work on a casual basis as a nurse in a children’s hospital. By chance she discovered through a colleague that her son was not living with her ex-husband as the court had intended. Out of concern for her son, she decided to seek the intervention of the court, even though she knew that she would never be awarded legal custody of her child. As matters turned out, the process took approximately two years, mainly because the ex-husband refused to appear in court when summonsed. When he finally did appear in court with the child (who was by then four years old), the appellant discovered that the child had been schooled to say that he was living happily with his father and his father’s second wife. As the court did not believe the appellant’s claims that the child had been sold, custody remained with the ex-husband. However, an amendment was made to allow the appellant access one day a week. The appellant was not, however, able to exercise her access rights as the ex-husband hid the child from her. Quite unexpectedly, in 1993 the appellant was visited by Mr and Mrs [X] who were concerned on a number of accounts. First they had been unaware of the true circumstances surrounding the birth and subsequent “adoption” of the child. Second, they had been unable to find their “son” after he had been taken away by the ex-husband. After combining forces, the appellant and Mr and Mrs [X] spent eight months searching before finally finding the child at the home of the ex-husband’s mother.
 The appellant and Mr and Mrs [X] then sought the intervention of the Karaj court on the grounds that the ex-husband had failed to comply with the earlier ruling on access. In July 1994 the court made an order which, while leaving formal custody of the child with the father, gave to the appellant the responsibility of caring for the child on a full-time basis. The order was made subject to the following conditions:
(a) The father was to have visiting rights.
(b) The appellant was not to move away from Karaj or to leave Iran.
(c) She was not to marry again. If she did, her son would have to be handed back to his father.
(d) Failure to abide by these conditions would not only result in the loss of the son, the appellant would also be sent to prison.
 The ex-husband was extremely angry that the appellant had won possession of their son, particularly given his position in Sepah. The court order was an embarrassment to him both as a man and as an official. He threatened the appellant that the child would not be with her for long. She feared that he would kidnap the child.
 By the time the appellant was reunited with her son in July 1994, he was almost six years old. Almost immediately the appellant was subjected to a renewed campaign of harassment by her ex-husband and by Pasdars acting on his behalf. There were repeated phone calls to her at her place of work. Men she did not know would ask her out and make indecent suggestions to her. At other times she would receive parcels such as perfume and flowers from people she did not know. In the street she would be stopped and harassed several times a week for allegedly failing to comply with the dress code. On other occasions, if she was travelling in a taxi, the vehicle would be stopped and questions asked as to why she was in a vehicle alone with a man. The Pasdars would say disgusting things to her and accuse her of having a sexual relationship with the driver. On other occasions the ex-husband appeared at her place of work dressed in uniform, thereby frightening the appellant, her employer and everyone else on the premises.
 In the period 1994 to 1995, in accordance with the court ruling, the appellant took her son to the mother-in-law’s house once a week so that the ex-husband could exercise his visiting rights. It transpired that he did not in fact have any interest in the child and that it was the mother-in-law who actually wanted to see the child.
 The appellant said that she was unable to complain to the authorities about her ex-husband’s conduct because he was too high an official in the Pasdaran and because she was a woman. When on occasion she had been taken to the Sepah office for questioning about alleged violations of the dress code or the like, she would ask the officers why they were harassing her. They would reply in offensive terms, telling her to shut up. They refused all her requests that she speak to senior officers. On one or two occasions while being questioned she had seen her ex-husband moving about the office.
 In September 1995, afraid for her safety and fearing that her son would be kidnapped by her ex-husband (he had begun visiting the daycare centre demanding to see the child), the appellant moved to Tehran at the suggestion of Mr and Mrs [X]. She got on well with them, as did her son who had been brought up to treat them as his mother and father. However, the appellant stayed in Tehran for only one year as she was afraid that her ex-husband would discover that she had left Karaj and demand that she surrender their child to him. Initially, to cover the fact that she was no longer living in Karaj, she would travel there from time to time to allow her ex-husband’s mother to see the child. On one of these visits the ex-husband’s mother told the appellant that the ex-husband had discovered that she was no longer living in Karaj. In November 1996 the appellant returned to Karaj with her son, living at her parents’ home.
 A short time after her return to Karaj the appellant renewed an acquaintance with a childhood friend and he proposed that they marry. The appellant was reluctant to accept as she did not wish to put at risk her de facto custody of her son. Eventually, however, the appellant relented and in April 1997 she entered into a five year temporary marriage with this man. As to the nature and significance of temporary marriage in Iranian society, see Parvin Paidar, Women and the Political Process in Twentieth-Century Iran (Cambridge University Press, 1995) 278-279; 284-286; Shahla Haeri, “Temporary Marriage: An Islamic Discourse on Female Sexuality in Iran” in Mahnaz Afkhami & Erika Friedl (eds), In The Eye of the Storm: Women in Post-Revolutionary Iran (Tauris, 1994) 98-114; Maryam Poya, Women, Work and Islamism: Ideology and Resistance in Iran (Zed Books, 1999) 68, 102-103; Research Directorate, Documentation, Information and Research Branch, Immigration and Refugee Board, Ottawa, Canada, Human Rights Brief: Women in the Islamic Republic of Iran (June 1994) para 2.6.1.
 After the marriage she moved back to Tehran hoping that her first husband would gradually lose interest in her and their son. However, a few months after her move to Tehran she was telephoned by her mother who reported that a few minutes earlier the first husband and several armed Pasdaran had burst into the family house in Karaj demanding to know the whereabouts of the appellant and her son. The mother had been beaten and, in fear of her life, had provided the appellant’s address in Tehran. The appellant, her second husband and her son immediately left Tehran for a town situated a short distance away. They did not return to the apartment. The decision made was that they would all leave Iran. The appellant already had a passport but it did not include her son. For an amendment to be lawfully made, she required the consent of her first husband. The problem was overcome when her second husband bribed an official in the passport office to add the child to the passport without evidence of consent. The appellant and her son immediately left Iran for Turkey. The plan was for her second husband to join her in a few weeks after he had liquidated their assets. However, from Turkey the appellant was unable to contact her second husband or her parents. In fear, she travelled to New Zealand with the help of an agent, arriving at Auckland Airport with her son on 4 March 1998. At the airport she sought refugee status. The statement taken from her at that time recorded the essential elements of her case.
 Since her arrival in New Zealand the appellant has learnt that her second husband has been harassed by her first husband. On one occasion he was beaten so badly that he was admitted to hospital for treatment. He was subsequently prosecuted in September 1998 for having false details added to the appellant’s passport and sentenced to six months imprisonment. He did not in fact serve any time as he was able to avoid the sentence by paying a large sum of money. Her first husband had also harassed members of the appellant’s family in Karaj, forcing them to change address. In addition, her mother has told her that the first husband has arranged for a warrant for the arrest of the appellant. The nature of the charge(s) is not known. Her second husband remains in Iran as his passport has been confiscated and he has been placed on a blacklist forbidding him from leaving the country.
 The appellant fears that should she return to Iran she will face imprisonment or other serious punishment for leaving Iran with her son on false documents and for kidnapping or removing the child from Iran. In addition, having breached the terms of the court order by leaving Karaj with her son, remarrying and leaving Iran itself, she will face the even greater penalty of being separated from her son. On top of that again, she fears that she will face further harassment from her first husband and his Pasdar colleagues.
 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.” In terms of Refugee Appeal No. 70074/96 Re ELLM (17 September 1996) 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 persecution?
ASSESSMENT OF THE APPELLANT’S CREDIBILITY
 The appellant impressed as a sincere, genuine and honest individual, albeit somewhat numbed by her experiences. In material respects her account of her marriage, divorce and long struggle to gain access to her son is corroborated by a number of court documents. The country information referred to earlier in this decision also provides independent support for her account.
 The appellant having been found to be a witness of the truth, her evidence is accepted in its entirety. This requires recognition of the fact that her first husband works in one of the most powerful of state departments (the Sepah Pasdaran) and has been able to influence events in his favour. Recent abuses of his Pasdar powers include the beating of the appellant’s mother to ascertain the appellant’s address in Tehran and the beating of the appellant’s second husband to extract a “confession” that a passport official had been bribed. The second husband received a seven month prison sentence while the first husband escaped even censure. These circumstances, along with the fact that a warrant has been issued for the arrest of the appellant, illustrate the degree to which the first husband is able to use his position in Sepah to manipulate the law to his personal advantage. There are no institutional checks in Iran on abuses of this kind.
 While the nature of the charge(s) is not known, it is our finding that there is a substantial chance that the appellant will face imprisonment or other serious punishment for having left Iran with her son on false documents. However, the consequences of her return go far beyond this. The appellant stands in breach all of the terms on which the care of her child was granted to her and will now forfeit her “rights” in that regard. Having seen and heard the appellant we are of the view that enforced separation from her son will have devastating consequences and will probably be the greatest penalty she faces. Quite apart from all of this there is a real danger of violence at the hands of her first husband and ongoing harassment of a substantial kind either from him or from his Pasdar colleagues. He has already forced her family members to change address to escape his attentions.
 For the avoidance of doubt we specifically find that:
(a) The fact that the appellant was, after a long and arduous struggle, able to gain access to her son (by then six years of age) does not permit the conclusion to be drawn that she was able to exercise at any meaningful level her fundamental rights as a woman or as the mother of her child.
(b) Given the institutionalised toleration, if not sanction, of domestic violence by the state authorities in Iran it would have been pointless for her, at any time from her marriage until her flight from Iran, to have complained about the actions of her first husband.
(c) In any event, the fact that her first husband held a position of influence and power within the Sepah Pasdaran foreclosed any theoretical opportunity she may have had to approach the state for assistance. It should not be forgotten that all requests that she be allowed to speak to senior officers were refused in an atmosphere of intimidation and violence.
(d) These same factors, coupled with her breaches of the “custody” order mean that should the appellant return to Iran she will not be able to access any state protection in relation to the actions of her first husband and his Pasdar colleagues.
 It is in the context of these findings that consideration must now be given to the questions of:
(a) Whether, if returned to Iran, the appellant faces persecution; and
(b) Whether the persecution anticipated by the appellant is for reason of her race, religion, nationality, membership of a particular social group or political opinion. This is the overriding issue. Unless a link or nexus can be established, the claim to refugee status must fail.
THE PERSECUTION ISSUE
Understanding the Meaning of Persecution
 Persecution is not defined by Article 1A(2) of the Refugee Convention and no purpose is served by attempting a definition of what is itself a definition. The interpretive approach to be followed is that prescribed by Article 31 of the Vienna Convention on the Law of Treaties, 1969. Article 31 (1) provides:A treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose. The three separate but related principles in this paragraph were identified by McHugh J in Applicant A v Minister for Immigration and Ethnic Affairs (1997) 190 CLR 225 at 252-253 (HCA). Brennan CJ, while dissenting as to the result, concurred at 231 with this aspect of the judgment of McHugh J, as did Gummow J (one of the majority) at 277. First, an interpretation must be in good faith, which flows directly from the rule pacta sunt servanda. Second, the ordinary meaning of the words of the treaty are presumed to be the authentic representation of the parties’ intentions. This principle has been described as the “very essence” of a textual approach to treaty interpretation. Third, the ordinary meaning of the words is not to be determined in a vacuum removed from the context of the treaty or its object or purpose. After referring to the controversy whether textual interpretation takes precedence over the object and purpose of the treaty, McHugh J preferred the ordered yet holistic approach taken by Zekia J in Golder v United Kingdom (1975) 1 EHRR 524, 544 (ECHR). That is, primacy is to be given to the written text of the Refugee Convention, but the context, object and purpose of the treaty must also be considered.
 This ordered yet holistic approach has been adopted by the Authority in Refugee Appeal No. 70366/96 Re C (22 September 1997) at 43-45;  4 HKC 236, 272-274.
 As a consequence, it is neither appropriate nor possible to distil the meaning of persecution by having resort to English and Australian dictionaries1. This can only lead to a sterile and mistaken interpretation of persecution. We refer by way of illustration of this point to the reference (with apparent approval) by Gummow J in Applicant A at 284 to a dictionary definition of “persecution” which asserted that it was the action of pursuing with enmity and malignity. This reference was subsequently treated by the Australian Refugee Review Tribunal as an authoritative statement as to how persecution was to be interpreted. The danger of this approach is twofold. First, it erroneously focuses on the intent of the persecutor rather than on the effect of the persecution on the victim. For a discussion of the difficulties this approach has caused in the United States see Deborah E Anker, Law of Asylum in the United States (3rd ed, 1999) 268-290. Second, and more relevant to the present context, it is an approach which lends itself to an unseemly ransacking of dictionaries for the mot juste appropriate to the case at hand. This does not assist in a principled analysis of the issues.
 Regrettably, when the issue resurfaced in the High Court of Australia in Chen Shi Hai v Minister for Immigration and Multicultural Affairs (2000) 170 ALR 553 (HCA) the opportunity was not taken to shift the interpretative approach adopted in Australia away from dictionaries and towards the more contextualised approach exemplified by Canada (Attorney General) v Ward  2 SCR 689 (SC:Can), namely that persecution may be defined as the sustained or systemic violation of human rights demonstrative of a failure of state protection. It is the latter approach which is now also favoured by the House of Lords. See Horvath v Secretary of State for the Home Department  3 WLR 379, 383B-H, 389A, 399H, 404F (Lords Hope, Browne-Wilkinson, Clyde and Hobhouse) (HL), a decision which should be read with R v Immigration Appeal Tribunal; Ex Parte Shah  2 AC 629, 644B-H, 648B, 651A, 652C, 653F, 658H (Lords Steyn, Hoffmann and Hutton) (HL). We are of the view that the dictionary approach exemplifies “the austerity of tabulated legalism” spoken of by Lord Wilberforce in Minister of Home Affairs v Fisher  AC319, 328H (PC). As McHugh J pointed out in Applicant A at 255-256:“Fourth, international treaties often fail to exhibit the precision of domestic legislation. This is the sometimes necessary price paid for multinational political comity. The lack of precision in treaties confirms the need to adopt interpretative principles, like those pronounced by Judge Zekia, which are founded on the view that treaties “cannot be expected to be applied with taut logical precision”....The phrase “a well-founded fear of being persecuted for reasons of ... membership of a particular social group” is a compound conception. It is therefore a mistake to isolate the elements of the definition, interpret them, and then ask whether the facts of the instant case are covered by the sum of those individual interpretations. Indeed, to ignore the totality of the words that define a refugee for the purposes of the Convention and the Act would be an error of law by virtue of a failure to construe the definition as a whole.” The New Zealand refugee jurisprudence as developed by this Authority has never employed the dictionary method of ascertaining the meaning of persecution. Instead it has followed Canada (Attorney General) v Ward  2 SCR 689. Delivering the judgment of the Supreme Court of Canada, La Forest J rightly recognised that fundamental to the Refugee Convention is the issue of state protection. The refugee scheme is surrogate or substitute protection, actuated only upon failure of national protection. See 709:“At the outset, it is useful to explore the rationale underlying the international refugee protection regime, for this permeates the interpretation of the various terms requiring examination. 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 p135.” The principle of surrogacy, long part of the Authority’s jurisprudence, has now been recognised also by the House of Lords in Horvath v Secretary of State for the Home Department  3 WLR 379, 383C, 389B, 404F (Lord Hope with whom Lords Browne-Wilkinson and Hobhouse agreed).
 Addressing the persecution element of the definition, the view taken by the Supreme Court of Canada in Ward at 733 was that underlying the Convention is the international community’s commitment to the assurance of basic human rights without discrimination. The passage which follows begins with the relevant recital from the Preamble to the Convention and concludes with the adoption of the analysis by Professor Hathaway that persecution may be defined as the sustained or systemic denial of basic human rights demonstrative of a failure of state protection:“Considering that the Charter of the United Nations and Universal Declarations of Human Rights approved on 10 December 1948 by the General Assembly have affirmed the principle that human beings shall enjoy fundamental rights and freedoms without discrimination.This theme outlines the boundaries of the objectives sought to be achieved and consented to by the delegates. It sets out, in a general fashion, the intention of the drafters and thereby provides an inherent limit to the cases embraced by the Convention. Hathaway, supra, at p.108, thus explains the impact of this general tone of the treaty on refugee law:The dominant view, however, is that refugee law ought to concern itself with actions which deny human dignity in any key way, and that the sustained or systemic denial of core human rights is the appropriate standard.This theme sets the boundaries for many of the elements of the definition of “Convention refugee”. “Persecution”, for example, undefined in the Convention has been ascribed the meaning of “sustained or systemic violation of basic human rights demonstrative of a failure of state protection”; see Hathaway, supra, at pp.104-105”. The consistently held view of the Authority has been that the principled approach of Ward to the interpretation of the persecution element of the refugee definition is to be preferred to the “dictionary” approach. The Authority has accordingly followed the example of the Supreme Court of Canada and adopted the formulation offered by Professor Hathaway in his seminal text, The Law of Refugee Status (1991) at 104,108 that refugee law ought to concern itself with actions which deny human dignity in any key way and that the sustained or systemic denial of core human rights is the appropriate standard. That is, core norms of international human rights law are relied on to define forms of serious harm within the scope of persecution. In his text at 106, Professor Hathaway initially identified the relevant core human rights as those contained in the so-called International Bill of Rights comprising the Universal Declaration of Human Rights, 1948 and by virtue of their almost universal accession, the International Covenant on Civil and Political Rights, 1966 and the International Covenant on Economic, Social and Cultural Rights, 1966. However, for the reasons Professor Hathaway has more recently and persuasively given, to the International Bill of Rights there should now be added the Convention on the Elimination of All Forms of Racial Discrimination, 1966 (CERD), the Convention on the Elimination of All Forms of Discrimination against Women, 1979 (CEDAW) and the Convention on the Rights of the Child, 1989 (CRC). See James C Hathaway, “The Relationship Between Human Rights and Refugee Law: What Refugee Law Judges Can Contribute” published in The Realities of Refugee Determination on the Eve of a New Millennium: The Role of the Judiciary (Proceedings of the 1998 Conference of the International Association of Refugee Law Judges, October 1998) 80, 85-90. We respectfully agree with this analysis and would only add that while the hierarchy of rights found in these instruments is not to be rigidly or mechanically applied, it does assist a principled analysis of the persecution issue.
 One further point which needs to be made is that the universality of the International Bill of Rights, CERD, CEDAW and the CRC will not permit social, cultural or religious practices in a country of origin from escaping assessment according to international human rights standards. This is fully explained in Refugee Appeal No. 2039/93 Re MN (12 February 1996) at 19-28. Speaking of these international instruments Professor Rosalyn Higgins in Problems and Process: International Law And How We Use It (Oxford University Press, 1994) 96-98 succinctly identified the underlying principle:“Human rights are rights held simply by virtue of being a human person. They are part and parcel of the integrity and dignity of the human being. They are thus rights that cannot be given or withdrawn at will by any domestic legal system ...It is sometimes suggested that there can be no fully universal concept of human rights, for it is necessary to take into account the diverse cultures and political systems of the world. In my view this is a point advanced mostly by states, and by liberal scholars anxious not to impose the Western view of things on others. It is rarely advanced by the oppressed, who are only too anxious to benefit from perceived universal standards. The non-universal, relativist view of human rights is in fact a very state-centred view and loses sight of the fact that human rights are human rights and not dependent on the fact that states, or groupings of states, may behave differently from each other so far as their politics, economic policy, and culture are concerned. I believe, profoundly in the universality of the human spirit. Individuals everywhere want the same essential things: to have sufficient food and shelter; to be able to speak freely; to practise their own religion or to abstain from religious belief; to feel that their person is not threatened by the state; to know that they will not be tortured, or detained without charge, and that, if charged, they will have a fair trial. I believe there is nothing in these aspirations that is dependent upon culture, or religion, or stage of development. They are as keenly felt by the African tribesman as by the European city-dweller, by the inhabitant of a Latin American shanty-town as by the resident of a Manhatten apartment.” Two other features of the New Zealand jurisprudence should be noted:
(a) It is recognised that various threats to human rights, in their cumulative effect, can deny human dignity in key ways and should properly be recognised as persecution for the purposes of the Convention: Refugee Appeal No. 2039/93 Re MN (12 February 1996) at 16. The need to recognise the cumulative effect of threats to human rights is particularly important in the context of refugee claims based on discrimination;
(b) The determination whether the treatment feared in any particular case amounts to persecution will involve normative judgments going beyond mere fact finding: Damouni v Minister for Immigration, Local Government and Ethnic Affairs (1989) 87 ALR 97, 101 (French J) adopted and applied Refugee Appeal No. 2039/93 Re MN (12 February 1996) at 15.
Discrimination as Persecution
 As this Authority recently emphasised in Refugee Appeal No. 71404/99 (29 October 1999) at para 65, it is important to bear in mind that discrimination per se is not enough to establish a case for refugee status. A distinction must be drawn between a breach of human rights and persecution. Not every breach of a refugee claimant’s human rights constitutes persecution. As pointed out by Professor Hathaway in The Law of Refugee Status (1991) at 103-104, the intention of the drafters was not to protect persons against any and all forms of even serious harm, but was rather to restrict refugee recognition to situations in which there was a risk of a type of injury that would be inconsistent with the basic duty of protection owed by a state to its own population:“As a holistic reading of the refugee definition demonstrates, the drafters were not concerned to respond to certain forms of harm per se, but were rather motivated to intervene only where the maltreatment anticipated was demonstrative of a breakdown of national protection” This does not diminish the point, however, that decision-makers should consciously strive both to recognise and to give proper weight to the impact of discriminatory measures on women: Refugee Appeal No. 1039/93 Re HBS and LBY (13 February 1995) 26.
Agents of Persecution
 While persecution may be defined as the sustained or systemic violation of basic human rights demonstrative of a failure of state protection, it must be made clear that 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.
 This point was emphatically made in Canada (Attorney General) v Ward  2 SCR 689, 709 (SC:Can):“The persecution alleged by the appellant emanates from non-state actors, the INLA; the Government of Ireland is in no way involved in it. This case, then, raises the question whether state involvement is a pre-requisite to “persecution” under the definition of “Convention refugee” in the Act. The precise issues are phrased differently by the parties, but can be summarized in the following fashion. First, is there a requirement that “persecution” emanate from the state? Second, does it matter whether the claim is based on the “unable” or “unwilling” branch of the definition? In my view, the answer to both these questions is no.” And later, at 716-717:“The international community was meant to be a forum of second resort for the persecuted, a ‘surrogate’, approachable upon 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, more widely, to provide refuge to those whose home state cannot or does not afford them protection from persecution. The former is, of course, comprised in the latter, but the drafters of the Convention had the latter, wider purpose in mind. The state’s inability to protect the individual from persecution founded on one of the enumerated grounds constitutes failure of local protection.I, therefore, conclude that persecution under the Convention includes situations where the state is not in strictness an accomplice to the persecution, but it is simply unable to protect its citizens.” As to the unable and unwilling issue, the Supreme Court of Canada at 719 rejected the suggestion that “unable” requires no state complicity, but that “unwilling” does. The Court found that the dichotomy is not supported by the text of the Convention or by the relevant authorities. The conclusion of the Court at 719 was that ineffective state protection is encompassed within the concept of “unable” and “unwilling” and at 720 that:“Whether the claimant is ‘unwilling’ or ‘unable’ to avail him-or-herself of the protection of a country of nationality, state complicity in the persecution is irrelevant. The distinction between these two branches of the ‘Convention refugee’ definition resides in the party’s precluding resort to state protection: in the case of ‘inability’, protection is denied to the claimant, whereas when the claimant is ‘unwilling’, he or she opts not to approach the state by reason of his or her fear on an enumerated basis. In either case, the state’s involvement in the persecution is not a necessary consideration. This factor is relevant, rather, in the determination of whether a fear of persecution exists.” The holding in Ward that state complicity in persecution is not a pre-requisite to a valid refugee claim has been expressly adopted and applied by this Authority in Refugee Appeal No. 2039/93 Re MN (12 February 1996) at 17-18. As that decision records, this Authority has from its first hearings in June 1991 (see Refugee Appeal No. 11/91 Re S (5 September 1991)) accepted that 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.
 As this Authority held in Refugee Appeal No. 71462/99 (27 September 1999) at para 47, the principle that state complicity in persecution is not a pre-requisite to a valid refugee claim flows from the language of Article 1A(2) itself and has been confirmed by the overwhelming trend of international refugee case law. The recent decisions of the House of Lords in R v Immigration Appeal Tribunal; Ex Parte Shah  2 AC 629 (HL) and Horvath v Secretary of State for the Home Department  3 WLR 379 (HL) have reinforced this trend.
The Standard of State Protection
 The refugee scheme is surrogate or substitute protection, actuated only upon failure of national protection. The question which arose in Horvath v Secretary of State for the Home Department  3 WLR 379 (HL) was the standard against which the sufficiency of state protection is to be measured where the agent of persecution is a non-state agent. Their Lordships unanimously rejected the submission that the level of protection provided by a state should be such as to reduce the risk to a refugee claimant to the point where the fear of persecution could be said to be no longer well-founded. The formula preferred by their Lordships was far less strict. In their opinion a refugee claimant who has a well-founded fear of persecution will not be recognised as a refugee if there is available in the home state a system for the protection of the citizen and a reasonable willingness by the state to operate it. See particularly Lord Clyde at 398D-E and 403F. Because this test expressly does not require a finding that the level of protection is such as to reduce the fear of persecution to below the well-founded standard, the English position is that an individual can be returned to his or her country of origin notwithstanding the fact that the person holds a well-founded fear of persecution for a Convention reason.
 With the greatest of respect, this interpretation of the Refugee Convention is at odds with the fundamental obligation of non-refoulement. Article 33(1) is explicit in prohibiting return in any manner to a country where the life or freedom of the refugee would be threatened for a Convention reason. This obligation cannot be avoided by a process of interpretation which measures the sufficiency of state protection not against the absence of a real risk of persecution, but against the availability of a system for the protection of the citizen and a reasonable willingness by the state to operate that system. The point which emerges from Ward is that the refugee inquiry is not an inquiry into blame. Rather the purpose of refugee law is to identify those who have a well-founded fear of persecution for a Convention reason. If the net result of a state’s “reasonable willingness” to operate a system for the protection of the citizen is that it is incapable of preventing a real chance of persecution of a particular individual, refugee status cannot be denied that individual. The persecuted clearly do not enjoy the protection of their country of origin. As La Forest J stated in Ward at 716:“The state’s inability to protect the individual from persecution founded on one of the enumerated grounds constitutes failure of local protection.” Addressing the specific issue of state willingness, La Forest J at 724 made a telling point which is not answered in Horvath:“Most states would be willing to attempt to protect when an objective assessment established that they are not able to do this effectively. Moreover, it would seem to defeat the purpose of international protection if a claimant would be required to risk his or her life seeking ineffective protection of a state, merely to demonstrate that ineffectiveness.” We would respectfully agree.
 In our view the proper approach to the question of state protection is to inquire 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. The duty of the state is not, however, to eliminate all risk of harm. This is the point made by Professor Hathaway in The Law of Refugee Status (1991) at 105 where he observes that we live in a highly imperfect world and that hardship and suffering remains very much part of the human condition for perhaps the majority of humankind.
 In summary, we are of the view that the decision in Horvath should not be followed in New Zealand on the issue of sufficiency of state protection. We do, however, agree with the majority of their Lordships (Lords Hope, Browne-Wilkinson, Clyde and Hobhouse) that in determining whether the particular facts establish persecution, the test is whether there is both a risk of serious harm and a failure of state protection. Adopting the formula employed in both Shah (653F) and Horvath (403B), Persecution = Serious Harm + The Failure of State Protection.
Is there a Presumption of State Protection?
 In Ward at 724-726 the Supreme Court of Canada addressed the issue how, in a practical sense, a refugee claimant proves a state’s inability to protect its nationals. The view taken was that “clear and convincing” confirmation of a state’s inability to protect must be provided as absent some evidence, nations should be presumed capable of protecting their citizens:“The issue that arises, then, is how, in a practical sense, a claimant makes proof of a state’s inability to protect its nationals as well as the reasonable nature of the claimant’s refusal actually to seek out this protection. On the facts of this case, proof on this point was unnecessary, as representatives of the state authorities conceded their inability to protect Ward. Where such an admission is not available, however, clear and convincing confirmation of a state’s inability to protect must be provided. For example, a claimant might advance testimony of similarly situated individuals let down by the state protection arrangement or the claimant’s testimony of past personal incidents in which state protection did not materialize. Absent some evidence, the claim should fail, as nations should be presumed capable of protecting their citizens. Security of nationals is, after all, the essence of sovereignty. Absent a situation of complete breakdown of state apparatus, such as that recognized in Lebanon in Zalzali, it should be assumed that the state is capable of protecting a claimant.” The Authority has for many years adopted and applied this reasoning, but only as an evidentiary rule.
 By way of conclusion on this section on agents of persecution, even if one was to categorise the actions of the appellant’s first husband as that of a non-state agent, the basic inquiry mandated by the Refugee Convention is not affected. In both state and non-state agent cases the inquiry is the same, that is, does the claimant have a well-founded fear of persecution for a Convention reason and is he or she unable, or owing to such fear, unwilling to avail him or herself of the protection of the country of nationality.
 Against this background it is possible to discuss the appellant’s case as it relates to the persecution issue.
Whether Appellant faces Persecution
 Applying the principles discussed, the question whether the appellant faces persecution on return to Iran is to be answered by determining whether the harm anticipated by her is serious harm and whether there will be an absence of state protection. A finding of persecution can only be made if the facts establish both serious harm and an absence of state protection. See Professor Hathaway, The Law of Refugee Status (1999) 125:“... in addition to identifying the human rights potentially at risk in the country of origin, a decision on whether or not an individual faces a risk of ‘persecution’ must also comprehend scrutiny of the state’s ability and willingness effectively to respond to that risk.” More recently in Shah (653F) and Horvath (403B) this approach has been expressed in the formula that: Persecution = Serious Harm + The Failure of State Protection.
The Serious Harm Issue
 The 1979 Iranian Constitution does not expressly relegate women to second-class status: Ann Elizabeth Mayer, Islam and Human Rights: Tradition and Politics (3rd ed, 1999) 113-114. However, the cumulative effect of the laws of Iran and of the so-called Islamic form of governance certainly produces that result. See for example the legal provisions concerning marriage, divorce and custody and the provisions of the Islamic Penal Code earlier referred to and the country information discussed in Refugee Appeal No. 2039/93 Re MN (12 February 1996).
 This state-legislated relegation of women to a substantially inferior status is in breach of fundamental human rights law which prohibits discrimination on the basis of gender. See particularly Articles 2, 3 and 26 of the International Covenant on Civil and Political Rights, 1966 (ICCPR) as well as the provisions of the Convention on the Elimination of All Forms of Discrimination against Women, 1979 (CEDAW), particularly Articles 2, 3, 15 and 16.
 Given that so much of the appellant’s case relates to the issues of marriage, divorce and custody, it is appropriate to record that Article 23(4) of the ICCPR specifically recognises the principle of equality in these spheres:“States Parties to the present Covenant shall take appropriate steps to ensure equality of rights and responsibilities of spouses as to marriage, during marriage and at its dissolution. In the case of dissolution, provision shall be made for the necessary protection of any children.” Articles 15 and 16 of CEDAW are even more explicit in this regard and no doubt other provisions of both CEDAW and of the ICCPR could be identified as further indicators of the degree to which human rights standards are breached by the Islamic Republic of Iran generally, and in relation to women in particular. Articles 15 and 16 provide:Article 151. States Parties shall accord to women equality with men before the law.Article 16
2. States Parties shall accord to women, in civil matters, a legal capacity identical to that of men and the same opportunities to exercise that capacity. In particular, they shall give women equal rights to conclude contracts and to administer property and shall treat them equally in all stages of procedure in courts and tribunals.
3. States Parties agree that all contracts and all other private instruments of any kind with a legal effect which is directed at restricting the legal capacity of women shall be deemed null and void.
4. States Parties shall accord to men and women the same rights with regard to the law relating to the movement of persons and the freedom to choose their residence and domicile.1. States Parties shall take all appropriate measures to eliminate discrimination against women in all matters relating to marriage and family relations and in particular shall ensure, on a basis of equality of men and women: Taking into account the cumulative effect of these breaches on the appellant, our conclusion is that the policy of gender discrimination and the enforcement of gender-based norms against women as a group in Iran is of a nature which permits a finding of persecution in the sense of a sustained or systemic violation of basic human rights.
(a) The same right to enter into marriage;
(b) The same right freely to choose a spouse and to enter into marriage only with their free and full consent;
(c) The same rights and responsibilities during marriage and at its dissolution;
(d) The same rights and responsibilities as parents, irrespective of their marital status, in matters relating to their children; in all cases the interests of the children shall be paramount;
(e) The same rights to decide freely and responsibly on the number and spacing of their children and to have access to the information, education and means to enable them to exercise these rights;
(f) The same rights and responsibilities with regard to guardianship, wardship, trusteeship and adoption of children, or similar institutions where these concepts exist in national legislation; in all cases the interests of the children shall be paramount;
(g) The same personal rights as husband and wife, including the right to choose a family name, a profession and an occupation;
(h) The same rights for both spouses in respect of the ownership, acquisition, management, administration, enjoyment and disposition of property, whether free of charge or for a valuable consideration.
2. The betrothal and the marriage of a child shall have no legal effect, and all necessary action, including legislation, shall be taken to specify a minimum age for marriage and to make the registration of marriages in an official registry compulsory.
 The appellant’s case, however, goes beyond the cataloguing of laws and actions which cause serious harm on the basis of gender. A compelling element in her case is the factor of her first husband. Through this vindictive and arbitrary member of the Sepah Pasdaran the appellant is at risk of both personal harm as well as state-sanctioned harm. He has in the past subjected her to physical and psychological violence, assaulted her mother in the presence of armed men and administered a severe beating to her second husband. He has instigated a vengeful prosecution against her knowing that as the principal witness and complainant he can ensure that she has no chance at all of a fair hearing - presuming against the evidence that there is such a chance. The harm sourced from the husband is compounded by the harm sourced from the state in the form of the severely discriminatory laws referred to earlier and which make it extremely difficult for women to obtain legal redress. We are satisfied that on these facts the appellant has established a well-founded fear of serious harm.
 By way of completeness we add that whatever justification there may be for prosecuting her for illegally taking her son from Iran, the “punishment” she will receive from both her husband and from the state (including permanent separation from her son) will be grotesquely disproportionate to such offence as she may be charged with. The facts do not establish legitimate prosecution.
The Failure of State Protection Issue
 Little needs to be said on this issue. The evidence establishes that the state itself has put in place the very legislative framework which to a large measure is the source of the serious harm faced by the appellant. In addition, the state itself condones, if not encourages the “private” or domestic violence which comprises the balance of the serious harm faced by the appellant. On any view of the facts, the state will fail to protect her should she return to Iran.
Conclusion on the Persecution Issue
 It follows from our findings that should the appellant return to Iran she will face both serious harm and a failure of state protection. The persecution element of the refugee definition has accordingly been established.
 It is now possible to turn to the issue whether this denial of protection will be for reason of the appellant’s race, religion, nationality, membership of a particular social group or political opinion.
THE CONVENTION GROUNDS
 A person who has a well-founded fear of persecution can only access the surrogate protection of the Refugee Convention if he or she can show that the anticipated persecution is for reason of one of the five grounds recognised by the Convention, namely race, religion, nationality, membership of a particular social group or political opinion. While the meaning of each of these grounds is a question of law, it is a question of fact whether the required nexus has been established in any particular case.
Race and Nationality
 The grounds of race and nationality clearly have no application in the present case.
Religion and Political Opinion
 Awareness is required that the Convention grounds of religion and political opinion will often have particular relevance in the context of a theocratic regime such as Iran. This is explained in Refugee Appeal No. 2039/93 Re MN (12 February 1996) at 28-31; 41-46, a case which also involved a refugee claim by a woman from Iran. In holding that her fear of persecution was for reason of religion and political opinion, the Authority relied (inter alia) on the following evidence:“It has been noted by Linda Cipriani in "Gender and Persecution: Protecting Women Under International Refugee Law" (1993) 7 Geo. Immigra. L.J. 511, 514 that the status of women in the Qur'an is exemplified by such passages as:"Men are the managers of the affairs of women because Allah has made the one superior to the other and because men spend of their wealth on women .... As for those women whose defiance you have cause to fear, admonish them, and keep them apart from your bed and beat them."The extrapolation of dictates of this nature is explained by Shahrzad Mojab in “Women from Iran”, op cit 14:“While inequality between women and men exists in all societies and in all religious traditions, it is important to realise that Islamic teachings on the physical, emotional and intellectual inferiority of women are no longer filed away in the Holy Book Koran or in other religious texts. We must remember that the Islamic principle of inferiority of women is now the basis of the policy of a despotic state that uses extreme forms of violence in order to regulate male/female relations on the basis of Islamic dogmas. The Islamic state uses without any restraint the enormous state power in order to regulate the life of women from the moment they are born to the last stage in the burial ceremonies. Every moment in the life of women is regulated in one way or another by the powerful state machinery.”...The point made by Ann Mayer in Islam and Human Rights: Tradition and Politics (2nd ed, 1995) 69 is that even at the more abstract constitutional level:"In the Iranian Constitution, Islam is conceived of, not as offering the basis for protecting rights, but solely as the basis for limiting or denying the rights that people could claim under secular constitutional principals or international human rights standards."And at op cit 70, she adds:"Iran's post revolutionary ruling elite became dominated by conservative clerics, who were eager to exploit their newly won political power to achieve clerical domination of society and whose lack of sympathy for human rights principles has been abundantly manifested by subsequent events. Both the conduct and the statements of the clerical members of Iran's ruling elite reveal that they see the official version of Islam as a tool of reverse social engineering for combatting the social changes that have accompanied Iran's modernization. In such circumstances, it is natural that Islamic qualifications of human rights embody retrograde interpretations of shari'a requirements and are used as instruments of repressive government policies."As observed by Fred Halliday in "Relativism and Universalism in Human Rights: the Case of the Islamic Middle East" Political Studies (1995) 43, 152, 161:"... what we have, behind claims to transhistorical and divinely sanctioned legitimacy, are projects for the acquisition and maintenance of political power in the late twentieth century."To similar effect, see the following passages from Reza Afshari, "An Essay on Islamic Cultural Relativism and the Discourse of Human Rights" Human Rights Quarterly 16 (1994) 235, 249:"For a Muslim country, as for all complex state societies, the most pressing human rights issue is not local cultural preferences or religious-cultural authenticity; it is the protection of individuals from a state that violates human rights, regardless of its cultural-ideological facade.... Under Islamist dictatorship in Iran, the drive for Islamization of culture has become a considerable smokescreen for the exclusionary (political and economic) strategy of the state."And at op cit 255:"Martha Nussbaum argues that 'one of the most central capabilities, without which one cannot consider a life as a human life, is 'the capability of choosing itself, which is made among the most fundamental elements of the human essence'. The state should not block access to information that enable citizens to examine alternative ways of life and to make informed choices among the possibilities open to them. Nor should it permit organized political groups to engage in violent actions which eliminate people's options. The angry young men of fundamentalism have turned the streets of Tehran into a veritable cultural war zone, harassing and attacking women in modern dresses. The situation in Cairo is only less drastic. Thus, women are at the centre of the debate over culture and authenticity. They are also central to the issue of the control and exercise of power in Muslim societies."These comments are particularly relevant given the earlier discussion in this decision of state protection, that is, the significance of the inability of the state, or its unwillingness, to afford adequate protection to individuals. The paper by Reza Afshari is also valuable for highlighting the political need in Iran for the continuing imperative to fulfill "Islamic" principles on women. The following quote is taken from op cit 257-258: As observed by Ann Elizabeth Mayer in Islam and Human Rights: Tradition and Politics (2nd ed, 1995) at 112:"The hegemony of the dominant Islamist polity places women at the centre of Islamist discourse. Manipulating the sacred symbolism, the Islamist discourse claims an extrapolitical sanctity, while its consecrated language empowers its speakers. I am referring to the Foucauldian sense of discourse, designating a materialization of language through which Islamist views have been positioned in absolute opposition to those of secularists. The Iranian Islamists trumpet their political legitimacy, playing Islamic tones no one else is said to be qualified to play. Their religious idiom is no longer as effective as when it was delivered from outside as a language of political dissent and discontent. In today's political marketplace, however, the issues for which the Islamist discourse could claim exclusive Islamist priorities and solutions have been diminished, as the hard-pressed ruling elite falls back on the familiar secular formulas of the past for administration and economic growth. As the state's dominant discourse, Islamism is left with only one major mark of distinction: its drive against moral impurities resulting from the 'un-Islamic' appearance and behaviour of women in public. Being placed on the edge of what seems to be the sole slippery terrain for deviation from the true path of Islam, women have become the raison d'être of Islamism, the reluctant bestowers of its legitimacy. The sight of a bare-headed woman is a challenge to that legitimacy. Other vices like alcohol, Western and Iranian popular music, videos, and films are produced locally, smuggled, distributed, and consumed in homes whose morality the Islamic Republic has failed to 'Islamitize'. Unlike women's dress, these are hidden vices in which men, Islamist or otherwise, partake without feeling threatened in their public domination.After more than a decade, Iran's Islamist regime has failed to bring about any fundamental change in the economic direction, in Iran's entrammelment in the global economy, and in its position as consumer of the developed world's goods and services. The new rulers have learned that there is no Islamic way to build and run an industrial plant, produce goods and services, or create monetary stability. What has remained of the Islamic version of the dependency theory is a skeleton of cultural rhetoric that will increasingly sound cranky to the swelling middle and lower-middle classes. As the Islamists have increasingly entangled themselves in other discourses (capitalist market economy, state capitalism, socialism, etc), over which they can lay no exclusive, divinely-sanctioned claims, they will feel compelled to wage their power struggle in cultural terms. Thus, women have become the signifier of the anti-Western credentials of the Islamists, the veiled ones as a positive testimony and the badly-veiled ones as the evidence for Islamist vigilance against the collaborators of the West. At the same time, reinforcing the hijab is not only a vocation for Islamist men who are sent out to the streets in sorties but an avocation for some of them who seem to enjoy the new opportunity to abuse women in 'bad' hijab and gain satisfaction for their frustrated sexual fantasies.""The record ... overwhelmingly establishes that Islamic principles, Islamic law, and Islamic morality have been interpreted in Iran to justify depriving women of any semblance of equality with men, subjecting them to a wide range of discriminatory laws and treatment, and effectively confining them to serving their husbands, performing domestic tasks, and bearing and raising children." Applying this evidence we find, subject to the appellant being able to satisfy the nexus requirement discussed below, that the religion and political opinion Convention grounds are directly applicable to her case.
 We turn to examine the only remaining Convention ground, namely membership of a particular social group.
PARTICULAR SOCIAL GROUP
 While on one view the holding we have just made as to the religion and political opinion grounds makes it unnecessary to examine the social group category, the point which must be made is that it is possible for Convention grounds to overlap. Because there is an overriding need to establish a nexus between the Convention ground and the anticipated serious harm, it is best to identify the principal or strongest ground in relation to which the “for reason of” inquiry is to be conducted.
 In the present context our view is that while the Iranian laws earlier discussed are designed, with supposed Islamic justification, to maintain political power, the overarching characteristic of the disenfranchised is their gender, that is the fact that they are women. This leads to the question whether Iranian women are a particular social group as that term is understood in Article 1A(2) of the Refugee Convention.
 The leading New Zealand case on the social group category is Refugee Appeal No. 1312/93 Re GJ (30 August 1995);  INLR 387, a decision cited with approval by Lord Steyn (Lords Hope and Hutton agreeing) in R v Immigration Appeal Tribunal; Ex Parte Shah  2 AC 629, 643E, 644G-H (HL).
 As indicated, the social group ground has been interpreted in recent years by the highest courts of Canada, Australia and the United Kingdom in Canada (Attorney General) v Ward  2 SCR 689 (SC:Can); Applicant A v Minister for Immigration and Ethnic Affairs (1997) 190 CLR 225 (HCA) and R v Immigration Appeal Tribunal; Ex parte Shah  2 AC 629 (HL). A large measure of consensus has emerged.
 First, the ambit of this element of the definition must be evaluated on the basis of the basic principles underlying the Refugee Convention. International refugee law was meant to serve as a “substitute” for national protection where the latter was not provided. The Convention has built-in limitations to the obligations of signatory states. These restricting mechanisms reflect the fact that the international community did not intend to offer a haven for all suffering individuals: Ward 731-732. The following passage is at 732:“... the drafters of the Convention limited the included bases for a well-founded fear of persecution to ‘race, religion, nationality, membership in a particular social group or political opinion’. Although the delegates inserted the social group category in order to cover any possible lacuna left by the other four groups, this does not necessarily lead to the conclusion that any association bound by some common thread is included. If this were the case, the enumeration of these bases would have been superfluous; the definition of ‘refugee’ could have been limited to individuals who have a well-founded fear of persecution without more. The drafter’s decision to list these bases was intended to function as another built-in limitation to the obligations of signatory states.” See also Applicant A at 247-248, 274, 283 and Shah at 638G-639D, 658H.
 Second, the particular social group category is limited by anti-discrimination notions inherent in civil and political rights: Ward 733, 739. Underlying the Convention is the international community’s commitment to the assurance of basic human rights without discrimination: Ward 733. This theme outlines the boundaries of the objectives sought to be achieved and consented to by the delegates who negotiated the terms of the Convention. It sets out, in a general fashion, the intention of the drafters and thereby provides an inherent limit to the cases embraced by the Convention. In distilling the contents of the head of “particular social group” therefore, it is appropriate to find inspiration in discrimination concepts. The manner in which groups are distinguished for the purposes of discrimination law can be appropriately imported into this area of refugee law: Ward 735. In short, the meaning assigned to “particular social group” should take into account the general underlying themes of the defence of human rights and anti-discrimination that form the basis for the international refugee protection initiative: Ward 739. See also Applicant A at 232 & 257 and Shah at 639C-D, 651A-D, 656E, 658H.
 Third, the ejusdem generis approach developed by the US Board of Immigration Appeals in Re Acosta 19 I & N, Dec. 211, 233 (BIA 1985) provides a good working rule in that it properly recognises that the persecution for reason of membership of a particular social group means persecution that is directed toward an individual who is a member of a group of persons all of whom share a common immutable characteristic. That characteristic must be either beyond the power of an individual to change, or so fundamental to individual identity or conscience that it ought not be required to be changed. What is excluded by this definition are groups defined by a characteristic which is changeable or from which disassociation is possible, so long as neither option requires renunciation of basic human rights: Ward at 736-737. See also Shah at 643C & 644D, 651E, 656F & 658E, 658H.
 Fourth, while the social group ground is an open-ended category which does not admit of a finite list of applications, three possible categories can be identified (Ward 739):
(a) Groups defined by an innate or unchangeable characteristic;
(b) Groups whose members voluntarily associate for reasons so fundamental to their human dignity that they should not be forced to forsake the association; and
(c) Groups associated by a former voluntary status, unalterable due to its historical permanence.
 The first category would embrace individuals fearing persecution on such bases as gender, linguistic background and sexual orientation, while the second would encompass, for example, human rights activists. The third branch is included more because of historical intentions, although it is also relevant to the anti-discrimination influences, in that one’s past is an immutable part of the person.
 Fifth there is a limitation involved in the words “particular social group”. Membership of a particular social group is one of only five categories. It is not an all encompassing category. Not every association bound by a common thread is included: Ward at 728-232, Applicant A at 242, 260 and Shah at 643B-C, 656D, 658H.
 Sixth, there is a general principle that there can only be a particular social group if the group exists independently of, and is not defined by, the persecution. Nevertheless, while persecutory conduct cannot define the social group, the actions of the persecutors may serve to identify or even cause the creation of a particular social group in society: Ward at 729, Applicant A at 242, 263-264, 285-286 and Shah at 639G-H, 645E, 656G, 658H, 662B.
 Seventh, cohesiveness is not a requirement for the existence of a particular social group. While cohesiveness may be helpful in proving the existence of a social group, the meaning of “particular social group” should not be limited by requiring cohesiveness: Ward at 739; Shah at 642A-643G, 651G, 657F, 658H, 661D.
 All of these principles are well established in the Authority’s social group jurisprudence. See particularly Refugee Appeal No. 1312/93 Re GJ (30 August 1995);  INLR 387 and Refugee Appeal No. 2039/93 Re MN (12 February 1996).
 What the Authority has stressed is the need for members of the particular social group to share an internal defining characteristic. It is not every group in society which is a particular social group for the purposes of the Refugee Convention. The following quote is taken from Refugee Appeal No. 1312/93 Re GJ at 56-57;  INLR 387, 422:“The mere fact that a person fears persecution by reason of a characteristic that he or she has in common with another person who also fears persecution, does not establish that the two are members of a particular social group for the purpose of the Convention.Herein lies the significance of the interpretative approach to the Refugee Convention discussed at length earlier in this decision and which recognises that the grounds of race, religion, nationality and political opinion focus on the claimant’s civil and political rights. The Acosta ejusdem generis interpretation of “particular social group” firmly weds the social group category to the principle of the avoidance of civil and political discrimination. In this way, the potential breadth of the social group category is purposefully restricted to claimants who can establish a nexus between who they are or what they believe and the risk of serious harm: Ward 738-739; Hathaway, The Law of Refugee Status (1989) 137. For the nexus criterion to be satisfied, there must be an internal defining characteristic shared by members of the particular social group. In the Acosta formulation, this occurs when the members of the group share a characteristic that is beyond their power to change, or when the shared characteristic is so fundamental to their identity or conscience that it ought not be required to be changed. In the very similar Ward formulation, the nexus criterion is satisfied where there is a shared defining characteristic that is either innate or unchangeable, or if voluntary association is involved, where that association is for reasons so fundamental to the human dignity of members of the group that they should not be forced to forsake the association.In this way, recognition is given to the principle that refugee law ought to concern itself with actions which deny human dignity in any key way: Hathaway op cit 108 approved in Ward at 733.” To similar effect see Applicant A at 264 per McHugh J:“The notion of persecution for reasons of membership of a particular social group implies that the group must be identifiable as a social unit. Only in the “particular social group” category is the notion of “membership” expressly mentioned. The use of that term in conjunction with “particular social group” connotes persons who are defined as a distinct social group by reason of some characteristic, attribute, activity, belief, interest or goal that unites them. If the group is perceived by people in the relevant country as a particular social group, it will usually but not always be the case that they are members of such a group. Without some form of internal linking or unity of characteristics, attributes, activities, beliefs, interests or goals, however, it is unlikely that a collection of individuals will or can be perceived as being a particular social group”. As can be seen from these principles it is indisputable that gender can be the defining characteristic of a social group and that “women” may be a particular social group. Depending on the facts, it may be unnecessary to define the group any further as in “women in Iran” because the “in Iran” element goes not to the identification of the group but to the identification of those in the group who face a real risk of harm.
Whether the Appellant is a Member of a Particular Social Group
 It now remains for these principles to be applied to the facts as we have found them.
 For the reasons given, the evidence relating to Iran establishes that the overarching characteristic of those fundamentally disenfranchised and marginalised by the state is the fact that they are women. This is a shared, immutable, internal defining characteristic. Applying the principles identified, we find that the particular social group is therefore women.
 In so formulating the group our findings mirror those made by the majority in Shah. See especially Lord Steyn at 644E-F and Lord Hoffmann at 652C. We acknowledge that on one view, the group so defined may be seen as a large and general one. However, two points must be made. The size of the group cannot be a limiting factor given the breadth of application of the other four Convention categories. Second, our finding is country specific. Particular Islamic regimes such as Iran and Pakistan present an extreme picture of discrimination against women.
 However, whether the appellant is a member of a particular social group is not the same question as whether the anticipated harm in Iran will be for reason of her membership of that group.
THE NEXUS ISSUE
 As pointed out by Dawson J in Applicant A at 240, the words “for reasons of” require a causal nexus between actual or perceived membership of the particular social group and the well-founded fear of persecution. It is not sufficient that a person be a member of a particular social group and also have a well-founded fear of persecution. The persecution must be feared for reason of the person’s membership or perceived membership of the particular social group.
 Accepting as we do that Persecution = Serious Harm + The Failure of State Protection, the nexus between the Convention reason and the persecution can be provided either by the serious harm limb or by the failure of the state protection limb. This means that if a refugee claimant is at real risk of serious harm at the hands of a non-state agent (eg husband, partner or other non-state agent) for reasons unrelated to any of the Convention grounds, but the failure of state protection is for reason of a Convention ground, the nexus requirement is satisfied. Conversely, if the risk of harm by the non-state agent is Convention related, but the failure of state protection is not, the nexus requirement is still satisfied. In either case the persecution is for reason of the admitted Convention reason. This is because “persecution” is a construct of two separate but essential elements, namely risk of serious harm and failure of protection. Logically, if either of the two constitutive elements is “for reason of” a Convention ground, the summative construct is itself for reason of a Convention ground. See Shah 646C-D, 648C, 653E-G and 654D.
 It is therefore important in a case where there is more than one agent of persecution to examine separately, in relation to each agent, the cause of the risk of serious harm or, as the case may be, the failure of state protection, such failure being established if the anticipated response of the state does not bring the risk of harm to below a well-founded fear.
 In this regard the decision in Khawar v Minister for Immigration and Multicultural Affairs (2000) 168 ALR 190 (Branson J) is more helpful than the majority and minority opinions of the Board of Immigration Appeals in In Re R-A-Int. Dec. 3403 (BIA 1999) delivered on 11 June 1999. In our respectful view, neither of the opinions meaningfully grapple with the issues and we note that the decision has not escaped criticism: Deborah E Anker, Nancy Kelly & John Willshire-Carrera, “Defining ‘particular social group’ in terms of gender: the Shah decision and US law” 76 Interpreter Releases 1005 (July 2, 1999); Karen Musalo, “Matter of R-A-: an analysis of the decision and its implications” 76 Interpreter Releases 1177 (Aug 9, 1999). The more recent decision of In Re S-A- Int. Dec. 3433 (BIA June 27, 2000) (noted in 77 Interpreter Releases 860 (June 30, 2000)) similarly fails to recognise that state protection can be denied to a gender-based social group by reason of the gender of the members of that group. See further Deborah E Anker, Law of Asylum in the United States (3rd ed, 1999) 392. But while we have found Khawar helpful, we believe that the obiter observation of Branson J at para 37 (that the failure of state protection “is itself capable of amounting to persecution”) must be read in context. On its own, failure of state protection is not capable of amounting to persecution. This is because persecution is the sum of serious harm plus failure of protection. It is clear that this is in fact how Branson J intended to be understood.
 We do not in this decision have to decide what, in the refugee law context, is the appropriate causation test, an issue also left open by Lord Steyn in Shah at 646. In that case Lord Hoffmann at 654E (with whom Lord Hope at 655H agreed) rejected as an oversimplification the proposition that the requirement of causation could be satisfied by applying the “but for” test. Contrast the Australian case law discussed by Mark Leeming in “When is Persecution for a Convention Reason?” (2000) 7 AJ Admin L 100 and there is also the more recent discussion in Chen Shi Hai v Minister for Immigration and Multicultural Affairs (2000) 170 ALR 553, paras 24-37 and 66-74 (HCA).
Nexus: The First Husband
 As to the feared serious harm at the hands of the first husband, we find that no nexus has been established in relation to the religion, political opinion and social group grounds. It is artificial to say that the serious harm likely to be inflicted on the appellant by the first husband is for reason of the fact that she is a woman or for religious or political reasons. There are many causes of violence in the home: Radhika Coomaraswamy, Preliminary report of the Special Rapporteur on violence against women to the Commission on Human Rights E/CN.4/1995/42, 22 November 1994, para 119.2 The same is true for post-separation custody disputes. The fact that (as recognised by the Preamble to the General Assembly resolution adopting the Declaration on the Elimination of Violence against Women (A/RES/48/104, 20 December 1993))3 violence against women is a manifestation of historically unequal power relations between men and women does not explain the reason why this appellant is at risk of serious harm at the hands of her first husband.
 We turn finally to the issue of nexus in the context of state harm and state protection.
Nexus: The State
 Our earlier findings at paras 74 to 79 were that the state in Iran condones, if not actively encourages, non-state actors such as husbands or former husbands to cause serious harm to women. In relation to this risk of non-state harm there will be an undoubted failure of state protection. Addressing now the issue of serious harm at the hands of the state itself and the failure of state protection in that regard, the evidence clearly establishes that the appellant is at risk of serious harm at the hands of the state and because the state is totalitarian in nature, no “state protection” will be available to her.
 As to the final question of nexus, we find that the reason why the appellant is exposed to serious state harm and to a lack of state protection both from the husband and from the state itself is because she is a woman. The cloak under which this persecution will ostensibly take place will be religion. Given that Iran is a theocratic state, this means also that the persecution will be for reason of political opinion. But as we have stated before, the overarching reason why the appellant is at risk of persecution is because she is a woman. The social group category is therefore the primary Convention ground in relation to which a nexus has been established.
 In conclusion our finding is that while the serious harm faced by the appellant at the hands of her first husband is not for a Convention reason, the failure by the state to protect her from that harm is for the Convention reasons of membership of a particular social group, religion and political opinion. The appellant is entitled to refugee status on this basis alone. However, the further finding we make is that the serious harm which the state itself will inflict on the appellant and the reason for its failure to protect her from that harm will be for precisely the same Convention reasons. The appellant is entitled to refugee status on this basis as well.
 The appellant is a person who holds a well-founded fear of persecution for a Convention reason. Refugee status is granted. The appeal is allowed.
[Rodger Haines QC]
1. The Convention in any event provides that the English and French texts are equally authentic.
2. The text of the Preliminary Report is conveniently reproduced in United Nations, The United Nations and the Advancement of Women 1945-1996 (1995) at 528.
3. The text of the Declaration is conveniently reproduced in United Nations, The United Nations and the Advancement of Women 1945-1996 (1995) at 474.