Introduction
Women, girls, men and boys all suffer when exposed to the effects of conflict. Sometimes women and girls have similar experiences in these situations as men and boys. Many times, however, they have different experiences. They may be subjected to different violations because they are women and girls, or they may be subjected to the same types of violations as men and boys but experience or perceive these harms in a different manner. One obvious example is sexual violence committed during war; the United Nations reports that ‘[s]exual violence, and the long shadow of terror and trauma it casts, disproportionately affects women and girls.’Footnote 1 This different female experience stems from pervasive global gender inequality: around the world, women and girls tend to be poorer and receive less education and are often less mobile as a result of traditional family and care-giving responsibilities, all of which negatively compound their experiences during conflict.Footnote 2 While women and girls may have common experiences based on their gender, sometimes girls suffer additional targeted harm as a result of their young age. For example, girls forcibly recruited to serve as fighters may serve in combat (like boys) but may also be subjected to conjugal slavery (unlike boys).
When women and girls fleeing conflict seek asylum, are their gender-differentiated experiences recognized by decision-makers? It is not immediately obvious that they would be, given that neither the terms ‘sex’ nor ‘gender’ appear in the definition of ‘refugee’ set out in the 1951 Convention Relating to the Status of Refugees (1951 Convention), as amended by the 1967 Protocol Relating to the Status of Refugees (1967 Protocol).Footnote 3 However, over the past twenty years, there has been a significant focus at the international level on ensuring a gender-sensitive interpretation of the refugee definition. These efforts include UNHCR’s issuance of a number of groundbreaking guidance documentsFootnote 4 and state adoption of legislation including ‘sex’ or ‘gender’ in the list of 1951 Convention groundsFootnote 5 or of guidelines on female asylum-seekers.Footnote 6
Given this guidance, one would expect gender-sensitive determinations of asylum claims by women and girls fleeing conflict. However, numerous studies show that deep flaws still exist in the domestic consideration of refugee claims by women and girls and that both policy and practice need improvement.Footnote 7 To date, this analysis has not concentrated specifically on women and girls who have fled conflict. This chapter therefore aims to bring some focus to this subset of female refugee claimants.Footnote 8 It does so by analyzing a group of forty-six cases decided between 2004 and 2012 in Australia, Canada, New Zealand, the United Kingdom and the United States involving women and girls who seek asylum based on claims related to conflict. These cases come from various levels of the refugee determination process. In addition to this group of recent cases, this chapter also considers certain relevant earlier cases, as well as reports on and academic discussions of the female refugee experience that include contemplation of conflict-related claims.
The second section of this chapter begins by exploring when conflict-related ill-treatment has been recognized as persecution and when it has not. This section also examines whether ill-treatment in conflict can be considered to be targeted at individual women and girls. The third section examines the 1951 Convention grounds most often used in refugee claims made by women and girls fleeing conflict: membership of a particular social group, political opinion, race and religion. The fourth section discusses how lack of state protection has been considered in female conflict-related refugee claims. The fifth section discusses procedural and evidentiary problems that arise in such asylum claims, especially with respect to credibility gaps on sexual violence and lack of gender-specific country-of-origin information. The final section concludes with observations on the need for a deeper understanding of persecution, expanded conceptions of the 1951 Convention grounds as they relate to women and girls fleeing conflict and recognition that women and girls fleeing conflict face problems similar to those making peacetime-related claims and specific conflict-related evidentiary and procedural hurdles.
Persecution
The 1951 Convention requires that the refugee claimant possess a well-founded fear of a form of harm that qualifies as persecution.Footnote 9 The term ‘persecution’ is not defined in the 1951 Convention, though there is agreement that ‘a threat to life or freedom on account of race, religion, nationality, political opinion or membership of a particular social group’ and other serious violations of international human rights constitute persecution.Footnote 10 There are a number of gender-related forms of harm that are currently considered to fit within these parameters – such as rape, dowry-related violence, female genital mutilation, domestic violence and trafficking – because they inflict severe pain and suffering (both mental and physical), whether perpetrated by state or non-state actors.Footnote 11 However, only some gender-related forms of ill-treatment common in the context of conflict have been recognized as rising to the level of persecution. The conflict-related harm that is most well established as a form of gendered persecution is rape.
Rape in Conflict
Rape is committed during conflict for many reasons.Footnote 12 These reasons may range from opportunistic ‘sexual looting’, to strategic design, to a combination of the two.Footnote 13 These reasons may overlap and coexist in a given conflict and may also change over time. Rape can also be committed in many ways, such as brutal gang rapes, the insertion of various objects into victims’ genitalia, the raping of pregnant women and forced sexual intercourse between male and female civilian abductees.Footnote 14 Rape is an expression by the perpetrator of control and powerFootnote 15 and therefore is effective in not only physically and psychologically harming the victims but also in tearing apart social units (such as families and communities).Footnote 16 This is why rape has been acknowledged as a particularly effective tool of genocide,Footnote 17 as a crime against humanity (including the crime against humanity of persecution)Footnote 18 and as a war crime.Footnote 19 Rape has also been recognized as a human rights violation.Footnote 20
Given the serious harm created by rape, it is therefore not surprising that rape has been identified within international and domestic refugee law as a form of persecution. UNHCR has stated that ‘[t]here is no doubt that rape is an act which inflicts severe pain and suffering (both mental and physical) and which has been used as a form of persecution by States and non-State actors.’Footnote 21 Various country guidelines, directed at refugee claim adjudicators, also specify rape as a form of persecution.Footnote 22 Domestic refugee case law has also recognized rape in conflict as a form of persecution.Footnote 23 In my examination of twenty recent cases involving claims of rape (of the claimant or a family member) or fear of rape as persecution,Footnote 24 eight claims or appeals were accepted on the evidence of past rape and/or future feared rape.Footnote 25 For example, in a New Zealand case involving the Democratic Republic of the Congo (DRC), the female applicant’s husband was sexually assaulted in detention, her mother and sister were raped during a visit to their house by soldiers and other women in her house were sexually assaulted. This, and other treatment during conflict, was considered to amount to a well-founded fear of persecution.Footnote 26
However, it must be noted that several of the eight positive decisions presented troubling histories in the lower courts, only remedied on appeal. For example, in the Canadian case of Kika, the claimant had been raped by soldiers in the DRC in 2006.Footnote 27 Her claim was initially denied on the basis of lack of evidence of persecution but was ordered to be reassessed, as ‘the officer apparently did not consider the possibility that Ms. Kika had a gender-based claim for refugee protection as a result of her sexual assault in 2006.’Footnote 28 As in Kika, the claimant in the case of NS (United Kingdom) was originally denied asylum in part because the adjudicator had found that her rape occurred ‘because the assailant found her attractive, and therefore that the attack was a purely personal one, and no more than a common crime’.Footnote 29 On appeal, the Immigration Appeal Tribunal stated that this ‘finding was not based on the evidence before [the adjudicator]’.Footnote 30 The evidence was that the applicant’s husband had been detained by the militia of an Afghan warlord on suspicion of supporting the warlord’s enemy.Footnote 31 The warlord’s nephew took advantage of her husband’s detention and the applicant’s vulnerabilityFootnote 32 and demanded that the applicant become his fourth wife; when she refused, he raped, beat and threatened her with death.Footnote 33 The tribunal concluded that ‘to take as a wife, by force, the wife of one’s enemy, after first imprisoning him, is not an uncommon act in the course of war or other conflict, as an act of aggression against the enemy.’Footnote 34 The applicant was granted refugee status.Footnote 35
In another case, PS (United Kingdom), the applicant, a Tamil woman from Jaffna, had been raped on three occasions in her home by Sri Lankan soldiers.Footnote 36 The first time, she was raped by two soldiers.Footnote 37 Five days later, one of those soldiers returned, accompanied by a different soldier, and both of them raped her. A week or so after, these latter two soldiers returned, held her father at gunpoint so that he could witness the act and raped her again.Footnote 38 She subsequently tried to kill herself and failed and then discovered that she was pregnant.Footnote 39 The immigration judge classified these soldiers as ‘rogue’, comparing them to three civilian criminals, and concluded that these past rapes had no relevance to the potential for future persecution.Footnote 40 This decision was rejected on appeal: the soldiers were clearly able to act with impunity, given the repetition of the rapes, and therefore there was a real risk that she would again be targeted for rape by Sri Lankan soldiers in the vicinity.Footnote 41 In a similar decision, in the case of LM Congo (United Kingdom), the court accepted, on appeal, that the applicant, who was from the Republic of Congo, had been raped in 1997 and that she had ‘a real risk of something similar happening to her on return’.Footnote 42 Finally, in In re B (FC), the House of Lords considered an appeal by a female Kosovar Albanian applicant who had been raped in front of her husband, sons and twenty to thirty neighbours due to suspected involvement in the Kosovo Liberation Army. Baroness Hale of Richmond found that the Court of Appeal had failed to realize that ‘the persecution of Mrs. B was expressed in a different way from the persecution of her husband and sons’, through rape, and that rape may be a weapon or strategy of war.Footnote 43
These examples demonstrate that rape claims stemming from conflict face some common obstacles. The first obstacle is in the characterization of the rape. Some adjudicators – such as the initial adjudicator in NS (United Kingdom) – view sexual violence in conflict as a matter of personal sexual gratification rather than as a method of terrorizing, controlling or punishing civilians.Footnote 44 In the UK case of Najjemba, the court held that a woman from Northern Uganda who had suffered rape at the hands of government soldiers was the victim of ‘simple and dreadful lust’ and not persecution.Footnote 45 This was despite her connection to her son, who was suspected of involvement with a rebel group opposed to the Ugandan government and had ‘disappeared’.Footnote 46
A related concern is that those who commit sexual violence in conflict are sometimes viewed by adjudicators as having committed the act in their private capacity and are therefore analogous to common criminals rather than persecutors, which was the original reasoning in PS (United Kingdom) and the reasoning in the Canadian case of VA9-00148.Footnote 47 Both the ‘conflict rape as lust’ and the ‘conflict rape as a common crime’ approaches demonstrate a profound misunderstanding of the context of rape in conflict. As Macklin puts it, ‘some decision-makers have proven unable to grasp the nature of rape by State actors [and non-state fighters, too] as an integral and tactical part of the arsenal of weapons deployed to brutalise, dehumanise, and humiliate women and demoralise their kin and community.’Footnote 48 The characterization of rape as lust incorrectly creates the impression that sexuality, rather than an exercise of power and gender-based discrimination, is at play. It also completely ignores the coercion, overarching violence and impunity created by the conflict. In other words, even if the perpetrator’s motive happens to be entirely sexual, it does not follow that the perpetrator did not target the victim for persecutory reasons or that his conduct does not cause severe pain and suffering to the point of persecution. Similarly, the characterization of rape as a ‘common’ criminal activity instead of persecution assumes that the rape in conflict is somehow random or private – and therefore unconnected to the state and lacking the discriminatory or rights-violating character required by the refugee definition.Footnote 49 This is despite the fact that the rape occurs in the context of overarching violence and impunity granted by weak or misdirected state authority and enhanced by vulnerability and gender-based discrimination. Both types of (mis)characterizations of rape are not exclusive to refugee claims involving conflict,Footnote 50 but the fact that they occur means that decision-makers are failing to adequately consider the environment surrounding the rape. On a positive note, these characterizations were correctly dismissed in In re B (FC) and the UK case of N.Footnote 51
The second common obstacle identified in the cases is that, sometimes, adjudicators (and the applicant’s counsel) do not recognize the importance of considering the after-effects of past rape in order to consider the risk of future persecution.Footnote 52 These after-effects commonly include societal stigma, which considers raped women and girls to be somehow ‘tainted’.Footnote 53 This stigma is a form of discrimination, and it can create social, cultural and economic exclusion for the victim and her children and lead to increased vulnerability to sexual and other forms of violence, as well as death.Footnote 54 For example, the US case of Mambwe considered the claim of a young Angolan woman who had fled from National Union for the Total Independence of Angola (UNITA) forces as a child, was raped in a Zambian refugee camp (and gave birth as a result) and was kidnapped from Zambia by UNITA and repeatedly raped.Footnote 55 Her appeal was denied on the basis that the conflict in Angola had ended.Footnote 56 However, it appears that the court (and earlier decision-makers) did not consider the lasting stigma (and therefore discrimination) she would likely face in post-conflict Angola as a lone young woman with no relatives who had been a past victim of rape by rebels, raising a child conceived through rape.Footnote 57 However, in In re B (FC), Baroness Hale of Richmond states, ‘To suffer the insult and indignity of being regarded by one’s own community (in Mrs. B’s words) as “dirty like contaminated” because one has suffered the gross ill-treatment of a particularly brutal and dehumanizing rape directed against that very community is the sort of cumulative denial of human dignity which to my mind is quite capable of amounting to persecution’ as ‘the victim is punished again and again for something which was not only not her fault but was deliberately persecutory of her, her family and her community.’Footnote 58 She did note, however, that these issues should have been identified by earlier decision-makers but were overlooked.Footnote 59 An evaluation of the risk created by the after-effects of rape should be a standard consideration in conflict-related cases in order to better understand potential sources of future persecution.
Other common obstacles to conflict-related refugee claims based on rape – such as rape being considered to be part of indiscriminate conflict, rape being found not to relate to a 1951 Convention ground, rape evidence being considered as not credible and a lack of relevant country-of-origin information on rape in conflict – will be dealt with in the sections that follow.
Other Forms of Sexual Violence in Conflict
While case law and academic analysis to date have largely focused on rape as a form of persecution common in conflict, other forms of sexual violence may qualify as forms of persecution.Footnote 60 International criminal law provides some assistance in enumerating examples of such sexual violence. Sexual slavery, enforced prostitution, mutilation of sexual organs or breasts and forced nudity have all been recognized as forms of sexual violence amounting to crimes against humanity and/or war crimes.Footnote 61 Given their comparability to rape, it can be expected that the same obstacles outlined earlier would apply to other forms of sexual violence.
Other Forms of Gender-Related Ill-Treatment in Conflict
There are a number of gender-related forms of ill-treatment that occur in conflict that may qualify as persecution. These include conjugal slavery (also referred to as ‘forced marriage’ in conflict),Footnote 62 forced pregnancy,Footnote 63 forced sterilizationFootnote 64 and forced abortion/miscarriage.Footnote 65 The latter three violations are likely the easiest to understand as gender-related crimes because they are targeted at reproductive organs. Conjugal slavery, which is comprised of sexual slavery plus domestic slavery, is also clearly a gender-related crime: the perpetrators enforce a norm of ‘femaleness’ on the victims, expecting them to submit to sex and cook and clean on demand.
Other forms of gender-related persecution in conflict may be more difficult to identify for two reasons. Firstly, applicants may describe seemingly gender-neutral forms of ill-treatment, such as torture, enslavement and imprisonment, but underneath those descriptions may lie a gendered form of the harm particularly targeted at or affecting women and girls: torture through threatened or actual sexual touching or rape,Footnote 66 enslavement by forced domestic labourFootnote 67 and imprisonment by the opposing side in a conflict so as to create easily available female ‘entertainment’ for troops.Footnote 68 Another example is terrorizing civilians; this may be achieved through a variety of methods, including rape, sexual slavery and conjugal slavery.Footnote 69 In other words, while some ill-treatment will be gender related on its face (like rape), other ill-treatment may only be revealed as such after gathering further information from the applicant. In the cases studied for this chapter, there appeared to be missed chances to discover potential gendered ill-treatment when ill-treatment was simply described as ‘torture’ or ‘beatings’ without further explanation of how these acts were carried out.Footnote 70
Secondly, the way in which women view harm may be gender differentiated. For example, ‘preliminary empirical research suggests that loss of a child, separation from children, and witnessing harm to children or family members are particularly viewed by women as primary harms to the self, often as or more egregious than a severe violation of their own bodies’.Footnote 71 Since female applicants may experience various harms as intertwined, they may not be able to easily compartmentalize the harms they face into conflict- and non-conflict-related harms for the purposes of describing past and feared persecution.Footnote 72 Similarly, given their experience of persecution as a physical-social-economic whole, they may not see sexual violence as the sole or central part of their claim.Footnote 73 Thus, applicants should be given the opportunity to identify the full range of conflict-related harms from their perspective in order for the harms to be evaluated cumulatively as persecution. Such an approach is more likely to uncover a wider range of gender-related persecutory acts.
Both reasons require those working with refugee claimants to listen closely to the stories told, to consider whether what is said reveals potential gendered ill-treatment and to ask appropriate questions to gain further insight.Footnote 74 In the cases studied, adjudicators tended to focus on the most obvious individual gender-related violation when considering whether persecution is gender relatedFootnote 75 rather than evaluating the harms as a whole, although there were exceptions.Footnote 76 However, sometimes gender aspects were simply never raised or examined by the applicant (or her representative) or the decision-makers, or the adjudicator decided to focus on the non-gender-related aspects of the claim.Footnote 77 This has been identified as a particular problem in cases involving sexual violence – ‘a tendency among some asylum professionals to marginalize, trivialize or ignore accounts of rape’.Footnote 78
Indiscriminate versus Targeted Gender-Related Ill-Treatment
This section concludes by discussing a major obstacle for female refugee claimants fleeing conflict: the categorization of gender-related violence as part of the general indiscriminate consequences of conflict and not targeted at the claimant. This classification occurred in a significant number of the cases studied.Footnote 79 For example, in the Canadian case of VA9-00148, an applicant from the DRC described an attack in 2008 in which rebels came to her home demanding money because they knew that her family had a business.Footnote 80 Her family was beaten and tortured, and the applicant was raped.Footnote 81 The rebels stole $40,000 and abducted her husband and stepson – she has not seen them since.Footnote 82 The rebel violence subsequently increased, and her neighbour and her children were beheaded, while others had body parts amputated or were killed.Footnote 83 While the adjudicator found that a nexus to a 1951 Convention ground does exist due to the sexual violence, he characterized the beatings, rape and torture of her and her family as ‘localized crime’, and even if the rebels had targeted them, it was for money and not for any other reason.Footnote 84 This led to a decision that ‘[a]lthough conditions in the DRC certainly involve a degree of risk and violence’, the applicant and her family would not face a personalized risk to their lives if returned.Footnote 85 This decision is particularly striking because it lacks any consideration of the significant political and ethnic dimensions of the conflict in the DRC, including the political and ethnic choices by the various rebel groups of where, how and who they attack.Footnote 86 As Goodwin-Gill and McAdam note, ‘A closer look at the background to the conflict … and the ways in which it was fought, will often establish a link to the [1951] Convention.’Footnote 87
The classification of gender-related ill-treatment in conflict as indiscriminate leads to the assumption that the applicant was not personally targeted for past persecution and/or that she can be returned to her country of origin because any future risk she would face is a risk faced by everyone in that country. While the facts of every case are individual, these underlying assumptions can be legitimately questioned. On the issue of personal targeting, it is crucial that decision-makers examine both the narrow and the wider context in which the violations occurred. In the case of VA9-00148 outlined earlier, the narrow focus reveals a potential case of targeting: the applicant was raped during the rebel attack, while her husband and son were abducted.Footnote 88 Both forms of mistreatment are gendered: the applicant may have been targeted for rape because she is female,Footnote 89 while the men may have been targeted for abduction because they are men (perhaps to become forced fighters). Furthermore, if one examined the wider context of militia movements in the applicant’s area at the time of the attack,Footnote 90 one might be able to discern other cross-cutting ways in which the applicant and her family were targeted – perhaps due to the (presumed) ethnicity or (presumed) political affiliation of the individuals in that area. These narrow and wider inquiries into the conflict would also inform analysis of the risk of future targeting for gender-related persecution (including persecution of a different type than originally suffered).Footnote 91 These deeper inquiries – both gender sensitive and intersectional – are crucial to more accurately determining whether violence is indiscriminate or targeted.
Lessons Learned
During war, gender norms often take on even greater socio-political significance than during peacetime. For example, ‘the role of women in the biological and social reproduction of group identity places them in a position of particular vulnerability.’Footnote 92 Thus, gender-related acts in conflict, such as the rape of women and girls, often take on deeper meanings or have broader repercussions (for families, for communities), thereby creating differentiated experiences. However, the significance of these meanings or repercussions is ‘currently rarely recognized in the asylum determination process’.Footnote 93 This is evidenced by the common obstacles faced by women and girls in proving that they were persecuted in gender-related ways in the past and/or that they risk gender-related persecution in the future. Within rape claims, there are continuing difficulties with the incorrect characterization of rape in conflict as a ‘private’ act. Additionally, not enough attention is paid to the after-effects of past rape in creating future risks of persecution. While the cases studied did not shed much light on other types of ill-treatment in conflict that might be considered as gender-related persecution, there are clearly many more (less obvious) forms.
Another, rather significant obstacle is that a number of decision-makers classify gender-related violence as part of the general indiscriminate consequences of conflict. It appears that this is done without necessarily considering potential gender-related reasons for targeting (e.g. the various ways in which rape is used as a weapon of war) or the wider political and other dimensions of the conflict. If this contextual and gender-sensitive analysis is done, it is suggested that fewer cases of gender-related ill-treatment would be categorized as untargeted.Footnote 94
Refugee decision-makers must listen carefully for, and draw out, gender-related ill-treatment while at the same time respecting that women or girls may identify harm in gender-differentiated ways. Unfortunately, the analysis of persecution in the cases studied suggests that this careful listening is not always happening.
1951 Convention Grounds
Under the 1951 Convention, only those who can demonstrate a ‘well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion’ can qualify as refugees.Footnote 95 As gender is not explicitly listed as a persecutory ground, women and girls fleeing conflict for gender-related reasons must fit their claims within one of the other grounds.Footnote 96 There are many different ways in which gender relates to the 1951 Convention grounds. If the persecutor would not have persecuted the victim had the victim not been female, ‘then an inference may be drawn that one of the motivations for persecution was the victim’s gender.’Footnote 97 For example, a party to a conflict may impose specific forms of conformity on women and girls based on a particular ideological view of how they should act.Footnote 98 When the reason underlying the persecution is the victim’s gender, then ‘membership of a particular social group’ may be the best category.Footnote 99 Where the gender of the victim dictates the manner of persecution (i.e. the persecution is carried out in a gender-specific manner, such as through rape and other forms of sexual violence, forced marriage, forced abortion, forced sterilization or forced pregnancy) but is not necessarily the reason for the persecution itself, then other 1951 Convention grounds might be more applicable.Footnote 100
Membership of a Particular Social Group
In practice, claims by women and girls tend to be considered under – and, indeed, funnelled into – the category of ‘membership of a particular social group’ (MPSG).Footnote 101 The cases studied confirmed this. MPSG was the most common 1951 Convention ground, with political opinion, race and religion the next most common grounds (in that order). Thus, conflict-related cases appear to reflect the more general trend in female cases, which suffer from the disproportionate use of MPSG.Footnote 102
UNHCR has defined ‘particular social group’ as ‘a group of persons who share a common characteristic other than their risk of being persecuted, or who are perceived as a group by society. The characteristic will often be one which is innate, unchangeable, or which is otherwise fundamental to identity, conscience or the exercise of one’s human rights.’Footnote 103 Within the cases studied, there were examples in which ‘women’ within a particular country were identified as such a group,Footnote 104 for example, Iraqi,Footnote 105 AfghaniFootnote 106 and Somali women.Footnote 107 However, in other cases, this category was not considered to provide enough differentiation.Footnote 108 In these other cases, the group was defined both by gender and by other characteristics, for example, Tamil women whose husbands are missing or dead,Footnote 109 Afghan women and girls related to a particular male,Footnote 110 lone Somali Ashraf woman with children,Footnote 111 widows of former members of the Iraqi Ba’ath Party and lone women with children,Footnote 112 single Somali women with children with no clan or family protectionFootnote 113 and family of senior Iraqi government employees with the additional factors of being in a mixed Sunni/Shia marriage and having liberal views.Footnote 114
The cases studied demonstrate that there are differing approaches as to whether and when it is appropriate to adopt the broad category of ‘women’ as a particular social group in a given country in conflict. For example, the UK case of HM indicated that women in Somalia form a particular social group ‘not just because they are women, but because they are extensively discriminated against’.Footnote 115 However, in the UK case of HH & Others, a narrower approach was adopted, in order, it appears, to be able to accept one female claimant and exclude others. Thus, the analysis did not focus, as it had in HM, on the overarching situation of women in Somalia. Rather, the focus was on clans and sub-clans because ‘[o]n the evidence, being a woman, without more, is not a sufficient differentiator’ to place her at individualized risk on return to a city ‘which is in a situation of armed conflict’.Footnote 116 Thus, the successful claimant was classified as part of the social group ‘lone Ashraf woman with children’, while the analysis of the others focused on clan (and not gender).Footnote 117 In this case, the narrowness of the social group seemed to be a decision-making device rather than an analysis of intersectionality. Intersectionality, in and of itself, can be positive and necessary because it recognizes the lived realities of female members of a society – who are not only female but also of a particular age, religion, race and so on.Footnote 118 However, if one is able to establish that a woman is being persecuted because she is a woman, or for reasons of gender, then ‘women’ may be the more accurate particular social group.Footnote 119
A default to the MPSG category sometimes also means that the nature of the conflict from which the applicant is fleeing is not analysed at all or not in depth.Footnote 120 It is not clear from the cases why this is so, but it may be because many MPSG gender-related cases typically deal with ‘private’ harms in peacetime,Footnote 121 such as domestic violence, forced marriage and female genital mutilation.Footnote 122 As with these other sorts of harms, sometimes the MPSG analysis in conflict-related cases focused closely on the ill-treatment and less so on the wider (and more ‘public’) religious, national or political aspects of the conflict and their gendered components.Footnote 123 As a result, potential social groups or other applicable 1951 Convention grounds may be missed. There is also the concern that conflict-related cases raising non-typical gender issues (e.g. outside of the realm of sexual violence) may mistakenly be considered as not qualifying as MPSG. Within the cases studied, the decisions that did consider the nature of the conflict in some depth tended to be more thorough in their consideration of the various facets of MPSG and the other 1951 Convention grounds.Footnote 124
It is well accepted that the social group cannot be solely defined by the type of current persecution.Footnote 125 However, this does not mean that the form of persecution is irrelevant. Those who have suffered past gender-related persecution such as rape ‘are linked by an immutable characteristic which is at once independent of and the cause of their current ill-treatment’.Footnote 126 A characteristic or an attribute expressed visibly may reinforce a finding that the individual belongs to a particular social group, but it is not a pre-condition for recognition of the group, especially given that those targeted for persecution may take pains to remain as invisible as possible.Footnote 127 Thus, ‘persecutory action toward a group may be a relevant factor in determining the visibility of a group in a particular society.’Footnote 128 This may be of special assistance in cases dealing with gender-related harm in conflict because women and girls who suffered war-related sexual violence or were conjugal slaves (‘bush wives’) may be stigmatized within society and therefore become part of a visible group. That said, some adjudicators seem to struggle with drawing the line between when a proposed particular social group is defined improperly by reference to conflict-related persecution and when a past form of conflict-related persecution might be a relevant factor in creating visibility for the immutable characteristic.Footnote 129
Contrary to these concerns, MPSG may also be a good category in which to analyse the types of conflict-related social and cultural harms identified by many women as central to their persecution.Footnote 130 This is illustrated, for example, in cases identifying relational social groups, especially family.Footnote 131 That said, women and girls should not be essentialized as solely social and cultural beingsFootnote 132 – obviously their lives are complex, and their suffering in conflict is also complex. It is for this reason that it is also important to consider the applicability of other 1951 Convention grounds in conflict-related refugee claims by women and girls.Footnote 133
Political Opinion and the Remaining Convention Grounds
The 1951 Convention ground of political opinion is particularly useful in conflict-related claims. Nine of the cases studied seriously considered political opinion as an applicable 1951 Convention ground.Footnote 134 This ground captures different ways in which a woman or girl may have political opinion imputed to her by a party to the conflict. This may occur when the claimant worked with, or for, a political party or a politician in her country of origin. For example, in the case of LM Congo (United Kingdom), the applicant had served as secretary for youth and in other roles for an opposition group.Footnote 135 Persecutors may also impute political opinion based on familial relationships. In the Australian case of MZXQS, the applicants claimed a well-founded fear of persecution because of their link to their sister, a well-known Tamil opposition Member of Parliament in Sri Lanka representing an LTTE-controlled area.Footnote 136 The original tribunal considered this claim as falling under MPSG and dismissed the claim on this ground.Footnote 137 The court found that the tribunal had incorrectly characterized the claim, which was actually a claim of imputed political opinion on the basis of their relationship with their sister.Footnote 138 Another way in which political opinion may be imputed to a woman or a girl is for appearing to hold views different from those of the warring factions.Footnote 139 In the UK case of LM Iraq, the applicant was perceived by Iraqi militia as supporting or collaborating with the West in part because she was a high-profile working woman who did not wear the hijab at work.Footnote 140 In addition, political opinion has been deemed to be imputed based on racial or ethnic identity. In the Australian case of 1203764, the tribunal found that there was a real risk that the applicant, a Tamil from the north, would be sexually assaulted or otherwise harmed if she returned to Sri Lanka for reasons of the political opinion imputed to her (membership of or sympathy for the LTTE).Footnote 141
The applicant herself may not classify her actions as political. It is therefore important for refugee decision-makers to recognize when a claim is, in fact, based on political opinion by examining the applicant’s actions. For example, in the case of 73894 et al., the female applicant, who was from the DRC, indicated that she ‘has not been interested in politics’.Footnote 142 However, she and her husband disagreed with the government’s policy of persecuting Rwandan Tutsis, and the applicant therefore helped to shelter Rwandan Tutsis (while her husband helped them to flee).Footnote 143 The Refugee Status Appeals Authority characterized this as ‘an overt political act opposing the policies of the Kabila regime’.Footnote 144 As a result of her actions and those of her husband, her home was searched several times, and the female members of her family were raped and sexually abused.Footnote 145 Despite the conclusion that sheltering Rwandan Tutsis was a political act, the authority felt that the applicant’s case was derivative of her husband’s and based its grant of refugee status on imputed political opinion (due to her husband’s actions).Footnote 146
Women’s political activity during conflict may take forms different from that associated with male political activity – and the political activity by women must be recognized as such. In the US case of Lopez, the applicant joined the Colombian Liberal Party, providing humanitarian assistance to residents of poor communities and conducting seminars on the principles of the Liberal Movement.Footnote 147 She was subsequently attacked by the FARC in retaliation for these activities.Footnote 148 The immigration judge found that these activities were ‘community-based and not political in nature’, and therefore the 1951 Convention ground of political opinion did not apply.Footnote 149 This categorization of women’s political activities as something other than political, such as community work, is also recognized as a serious problem for non-conflict-related female refugee claims.Footnote 150
Finally, the cases demonstrated that in the context of racially or religiously motivated conflicts, the 1951 Convention grounds of race and religion are particularly helpful.Footnote 151 While the 1951 Convention ground of nationality was not represented in the cases studied, it also would be a useful ground in the context of nationality-driven conflicts.
Lessons Learned
This section has explored the hazards associated with over-reliance on MPSG in cases dealing with both gender and conflict, such as a tendency towards creating artificial subgroups of women rather than relying on ‘women’ generally.Footnote 152 Additionally, there appears to be a focus on the ‘private’ side of the gender-related harms to the detriment of an analysis of the nature of the conflict. It is important for adjudicators to avoid an automatic reliance on MPSG and instead consider the other 1951 Convention grounds, especially political opinion. Race, religion and nationality may also be useful grounds when considering cases stemming from racial, religious or nationality-driven conflicts. However, there are benefits to using MPSG, as it draws attention to social and cultural harms.
Lack of State Protection
When adjudicators evaluate the risk of future persecution in gender-related claims, one issue they assess is whether the applicant can benefit from state protection against the actions of non-state actors. The cases studied revealed some gender-sensitive analysis of whether or not state protection is available, especially consideration of the dangers faced by a lone woman (sometimes with children) returning to a conflict-ridden society rife with discrimination against women, in which women (and girls) may be common targets for gender-related violence (such as rape, forced prostitution and trafficking).Footnote 153 However, there were also disturbing examples of non-sensitivity. In the UK case of PS, the second immigration judge found that the Tamil applicant’s rapists, despite being Sri Lankan soldiers, were comparable to ‘three civilian criminals’.Footnote 154 The judge concluded that there was no threat of future persecution and that, in the event the applicant had difficulties from them again, she could seek state protection from the Sri Lankan government.Footnote 155 This was overturned on appeal, with the judges finding that ‘[t]he whole point was that, unlike ordinary criminals, the soldiers were in a position to commit and repeat their crime with no apparent prospect of detection or punishment.’Footnote 156 The lesson from this case is that the analysis of future risk must be undertaken in a gender-sensitive manner with a full appreciation of the nature of the conflict, including whether the state permits impunity for gender-related violations.
One key issue arising in the cases studied related to the impact the end of a conflict had on consideration of the risk of future persecution and state protection. This is demonstrated in the US case of Mambwe, in which the end of the civil war in Angola was considered to eliminate any future risk of persecution.Footnote 157 This was despite the applicant’s assertion that the civil war ‘was not put to rest’ by the peace accord and disarmament of UNITA.Footnote 158 Rather than consider persecutory risks in Angola facing young female rape victims of UNITA with a child by rape who have no relatives, the court instead only considered whether UNITA is still a military threat.Footnote 159 Thus, state protection from sources of persecution other than UNITA were not considered, even though the applicant is likely to face severe societal stigma from those on both sides of the conflict.Footnote 160 It is important to recognize that the timelines of persecution do not necessarily accord with the timelines of cease-fires or peace agreements. One court explained, ‘Regime changes may be less effective in protecting women from such dangers [as rape] than they are for men.’Footnote 161 The Canadian guidelines correctly state, ‘A change in country circumstances, generally viewed as a positive change, may have no impact, or even a negative impact, on a woman’s fear of gender-related persecution.’Footnote 162 Peace processes may marginalize women’s concerns and may not touch deep-seated discrimination directed against women and girls.Footnote 163 When considering risk of future persecution in cases where conflict has ceased, it is relevant for an adjudicator to consider whether conflict-related sexual violence has been addressed in any cease-fire or peace agreement in the country of origin.Footnote 164 If it has not, then this is a potential indicator of state unwillingness to counter sexual violence.
Procedural and Evidentiary Issues
Procedural and evidential barriers ‘often inhibit women’s access to the determination process and may serve to limit the quality of information gathered about the claim and, in turn, the quality of the decision-making process’.Footnote 165 This proved true in the cases reviewed. The most challenging issue arising in conflict- and gender-related claims appears to be lack of gender-sensitive country-of-origin information, followed closely by an inability to establish credibility.
Country-of-Origin Information
The lack of gender-sensitive country-of-origin information is an overarching problem affecting all gender-related claims,Footnote 166 but the problem seems to be compounded in conflict-related claims. Where there was adequate gender- and conflict-related country-of-origin information available to decision-makers, the analysis of the cases tended to be more thorough and sensitive.Footnote 167 Where such information seemed to be lacking, the analysis was less thorough and more speculative, and the female applicants had serious difficulties proving their cases.Footnote 168
Female claimants typically would benefit from pre-conflict information on the legal, political, social, cultural and economic position of women and girls and the consequences for non-adherence to socio-cultural gender norms; information on how these aspects have changed for women and girls during the conflict; the incidence and forms of reported violence (in both the private and public spheres) against women and girls pre-conflict and during the conflict; the protection available to them during the conflict or post-conflict; any penalties imposed on those who perpetrate the violence and detailed information about the nature of the conflict and the parties to the conflict.Footnote 169 It is not always possible to collect this information on countries at peace, but getting accurate, up-to-date information on the situation of women and girls during a conflict can be extremely difficult, and if it is collected, it likely reflects under-reporting and therefore underestimation.Footnote 170
The cases studied tended to rely on specific types of country-of-origin information, especially from UNHCR,Footnote 171 so it is crucial that UNHCR continue to provide as much guidance in this respect as possible.Footnote 172 Other sources included international NGOsFootnote 173 and certain UN reports.Footnote 174 There is certainly scope for improving country-of-origin information on both gender and conflict issues to include a wider range of UN documents, such as Security Council resolutions referring to gender-related ill-treatment,Footnote 175 reports of the UN Secretary General written pursuant to Security Council resolutions 1889 (and its indicators) and 1960Footnote 176 and other UN reports providing qualitative and quantitative information on women and girls in conflict settings. Finally, reports from international and domestic nongovernmental women’s organizations (including those located in the country of origin) should be considered.Footnote 177 All refugee-receiving countries should aim to systematically collect and make available to applicants and their representatives up-to-date and accurate information on the situation and experiences of women and girls, including in conflict.Footnote 178 Where there is a lack of information, decision-makers should be cautioned against drawing speculative conclusions or assuming lack of persecution.Footnote 179
Credibility
The second most challenging procedural barrier relates to credibility. Within the cases studied, a number of claims were not accepted due to rulings of lack of credibility, either at the initial stages or on appeal.Footnote 180 This was due to a number of factors, most often inconsistenciesFootnote 181 or perceived implausibilities in testimony, ‘incorrect’ demeanour (e.g. being matter-of-fact when the adjudicator expects an applicant to be distressed)Footnote 182 or lack of corroborative country-of-origin information. When found credible, it was often due to a combination of ‘correct’ demeanour,Footnote 183 relative consistency in the applicant’s storyFootnote 184 and corroborative country-of-origin information.Footnote 185 The cases mirror concerns expressed about similar experiences with claims by women and girls more generally.Footnote 186 For example, previous studies have shown that most female claimants are simply not believed at first instance.Footnote 187 This is due to many factors: a hostile environment negatively affecting how detailed the applicant can be in explaining her case,Footnote 188 undue concentration on perceived inconsistencies without consideration for the impact of trauma and dislocation on memory or for culturally different ways of expression,Footnote 189 disincentives for women and girls to reveal sexual violence (due to being traumatized, feelings of shame or fear of stigma) with late disclosure of sexual violence sometimes being held against the applicant,Footnote 190 difficulty in evidencing gender-specific forms of harm and the absence of state protectionFootnote 191 and incorrect assumptions about the meaning of an applicant’s demeanour.Footnote 192 Credibility findings are clearly affected by gender-insensitive refugee claim processes and procedures.
Lessons Learned
The cases revealed two major procedural problems facing women and girls making refugee claims based on a combination of gender- and conflict-related harms. The first challenge for these applicants was in accessing and presenting accurate and up-to-date country-of-origin information containing relevant facts about the conflict and its gender dimensions. The second difficulty was in establishing credibility in the claim procedure. These problems are not exclusive to conflict-related claims, but the fact that a claim involves conflict heightens these challenges.
Conclusions
Women and girls fleeing conflict clearly face a number of obstacles to presenting a successful claim for refugee status. In the second section, this chapter indicated that some forms of gender-related ill-treatment in conflict have been found to amount to persecution, especially rape. And yet many applicants have difficulty establishing conflict-related rape as persecution for two reasons. Firstly, some adjudicators incorrectly characterize rape in conflict as a ‘private’ act and therefore outside the realm of persecution, and secondly, not enough attention is paid by decision-makers to the after-effects of past rape in creating related yet different future risks of persecution. Another challenge relates to the current relatively narrow perception of what qualifies as a gender-related form of persecution in conflict. Rape and some other forms of sexual violence are recognized, but it is less common for decision-makers to recognize non-sexual but still gendered violations. A third obstacle is that some decision-makers classify gender-related violence as part of the general indiscriminate consequences of conflict and therefore not targeted enough to amount to past persecution or present a risk for future persecution. However, refugee decision-makers may be less likely to classify such violence as untargeted if they have an in-depth understanding of both the gendered nature of the conflict and the nature of gender-related discrimination before, during and after the conflict, provided through country-of-origin information. While it is not clear whether there are forms of gender-related ill-treatment in conflict that may properly be considered as indiscriminate, it is evident that more cases of such ill-treatment should be considered as targeted (rather than indiscriminate) than is currently the case.
The third section continued the discussion of obstacles, focusing on the 1951 Convention grounds. As with all gender-related claims, there is a tendency for adjudicators to rely on MPSG as the main ground for analysis. The use of MPSG can be acceptable, but it can also be problematic. For example, while some decision-makers have accepted that ‘women’ can be a valid particular social group, others create narrower, sometimes artificial, subgroups, which can distort the resulting analysis. As well, some adjudicators focus on the ‘private’ side of the gender-related harms to the detriment of an analysis of the nature of the conflict. These obstacles may be overcome in conflict-related cases with a more fulsome focus on the other 1951 Convention grounds, such as political opinion.
The fourth section considered obstacles arising with the consideration of whether or not there is state protection in the country of origin. There is some positive case law considering the risks to lone females returning to conflict-affected countries of origin. However, there are also cases in which the judges failed to understand the actual vulnerabilities of women and girls in relation to their own conflict-affected or post-conflict state. Additionally, some positive case law recognizes that a cease fire or peace agreement does not necessarily mean the end of gender-related persecution in a country of origin, but other case law demonstrates that some decision-makers do not pay enough attention to the actual post-conflict circumstances of women and girls. Women and girls fleeing conflict therefore face more difficulties than they should in demonstrating lack of state protection and risk for future persecution.
Finally, the last section examined procedural and evidentiary difficulties arising in conflict- and gender-related claims. Women and girls face significant hurdles in accessing and presenting accurate and up-to-date country-of-origin information containing relevant facts on the conflict and its gender dimensions; this has a significant impact on their ability to demonstrate persecution, a 1951 Convention ground, and lack of state protection and therefore their credibility. On the latter, some decision-makers took into account the effects of conflict-related trauma on memory or demeanour, but others unfortunately did not.
In conclusion, while there are welcome developments in international and domestic refugee law under which claims by women and girls fleeing conflict have been accepted, there is also significant room for improvement. There is a need for a deeper understanding of gender-related persecution and its future risks, such that seemingly indiscriminate and/or seemingly gender-neutral ill-treatment of women and girls is more correctly recognized as persecution. There is also a need for expanded conceptions of the 1951 Convention grounds as they relate to women and girls fleeing conflict. Lastly, while women and girls fleeing conflict face problems similar to those making peacetime-related claims, they may also face specific conflict-related evidentiary and credibility hurdles. There have been many recent studies outlining proposed improvements to domestic refugee determination processes involving either women and girls or gender-based claims more generally.Footnote 193 The hurdles identified in this chapter might be (partly) removed by implementing the recommendations in these studies, but it may be necessary for decision-makers to be provided with (more) conflict- and gender-specific guidance on persecution, applicable 1951 Convention grounds and state protection to assess country-of-origin information and credibility.
