Speaking the Unspeakable: Disclosures of Sexual and Gender-based Violence in Asylum Credibility Assessments

ABSTRACT This article explores the extent to which the human rights framework relating to sexual and gender-based violence (SGBV) is upheld in the Norwegian asylum system, by investigating if and how asylum bureaucrats enable the disclosure of SGBV and how such disclosure may impact the assessment of applicants’ credibility. Credibility assessment is important in deciding who is eligible for protection. Eliciting disclosures of SGBV in general is notoriously difficult, and SGBV allegations are often disbelieved. It is also well known that credibility assessment in the asylum system involves working with ambiguity and challenges, but this has rarely been explored with a particular focus on SGBV. Through an analysis of public case summaries and 18 semi-structured interviews with asylum caseworkers and key actors, drawing on Lipsky’s theory of street-level bureaucrats, we conclude that caseworkers’ use of coping mechanisms makes them reluctant to enable, or engage with, disclosures of SGBV, and that this may endanger human rights.


Introduction
Today, there are an unprecedented 89.3 million forcibly displaced people worldwide, categorised as either asylum-seekers, 1 refugees, or internally displaced persons. 2 Almost 50% of them are women, 3 some of whom have fled sexual and gender-based violence (SGBV) such as assault, harassment, and rape. 4 As most recently demonstrated by the widespread sexual violence against women and girls fleeing the war in Ukraine, 5 street-level bureaucrats in the asylum context and the decision-making mechanisms they rely on throughout the credibility assessment.
This study contributes to the emerging research on the role of street-level bureaucrats in the asylum process. Using the Norwegian asylum system and the Norwegian Directorate of Immigration (UDI, 'Utlendingsdirektoratet') as a case study, we interview 18 individuals with profound knowledge of the asylum decision-making process. Additionally, we analyse 85 public case summaries related to SGBV that were contested in 2015-2020. We explore: (1) whether the two coping mechanisms of categorisation and standardisation that Lipsky describes can be identified among the asylum caseworkers, and (2) how these coping mechanisms may impact whether and how disclosures of experienced SGBV by women applicants are enabled and responded to. Our results indicate that categorisation and standardisation impact the asylum process in ways that may compromise women asylum-seekers' potential entitlement to protection. This study thus makes a theoretical and empirical contribution to previous research by analysing unique data from a previously unexplored empirical context.

Previous studies on asylum processes and credibility assessments
States are responsible for ensuring the human rights of their citizens. When they are unable or unwilling to do so, refugees have a right to seek protection in other countries, and the principle of non-refoulement prohibits the return of an applicant to a territory where his or her life or freedom is threatened. 16 The 1951 Convention relating to the Status of Refugees (hereafter 'the Refugee Convention') states that a person can be considered a refugee if s/he has a well-founded fear of being persecuted because of his or her race, religion, nationality, membership of a particular social group or political opinion; and is unable or unwilling to avail him-or herself of the protection of that country, or to return there, for fear of persecution. (UN General Assembly; UNGA, 1951, article 1A(2)) Central to most literature on migration and asylum are the challenges that arise from treating refugees as a clear-cut category where you can distinguish those who deserve protection from those who do not. Most democratic states employ language that emphasises the value of respecting human rights by offering protection, while adhering to the principle of preserving sovereignty and border control. 17 Arguably, asylum decisionmaking is one of the more complex areas of bureaucracy in the modern state.
The credibility assessment is at the core of asylum decision-making. Because documentary evidence to substantiate the applicant's claim is often lacking, more weight is placed on testimonial evidence. 18 The UNHCR has cited credibility as established if the applicant's claim is 'coherent and plausible, not contradicting generally known facts, and therefore is, on balance, capable of being believed'. 19  evaluate the internal and external credibility of an asylum claim and the UNHCR promotes a harmonised approach to credibility assessment across the region. 20 Internal credibility pertains to the presentation of the applicant's narrative in terms of its 'level of detail, demeanour, coherence, evasiveness, plausibility and consistency'. 21 External credibility concerns whether it is consistent with external country information.
A wide array of disciplines, such as ethnology, anthropology, sociology, criminology, psychology, and law have established that determining credibility is extremely difficult. 22 It is recognised that credibility assessment offers myriad challenges and that vague definitions such as well-founded and plausible grant discretionary power to caseworkers. 23 Arguably, assessment relies on subjective and normative assumptions regarding what a credible account looks like. 24 The next section will highlight how the credibility assessment may offer some unique challenges in the context of women applicants with experiences of SGBV.

Previous studies on SGBV, gender, and migration
For a long time, the topics of women and gender were missing from research and policymaking on migration. Recently, women migrants and refugees have increasingly been identified as 'particularly vulnerable' to SGBV, experiencing gendered forms of violence throughout the continuum of their journeys: pre-departure, during, and after arriving in the host country. 25 Europe's increasingly strict migration and asylum policies might have exacerbated this, because journeys are becoming more dangerous, migrants and refugees spend more time in camps and detention facilities, and they must frequently rely on traffickers. 26 Keygnart et al. found that women applicants and refugees in Belgium and the Netherlands had experienced alarming rates of SGBV, particularly sexual violence. 27 In her research on women migrants in Morocco, Tyszler discovered that all those travelling by road had suffered sexual violence. 28 Freedman concludes that such a large number of women migrants and refugees have been raped or suffered violence at the hands of soldiers, other migrants, or civilians that SGBV must be considered systemic in this context. 29 In her study of women applicants in the UK, Sweden, and Denmark, Canning found that they all listed ongoing or previous instances of SGBV as a main 20 UNHCR, Beyond Proof: Credibility Assessment in the EU Asylum Systems [2013], <https://www.unhcr.org/protection/ operations/51a8a08a9/full-report-beyond-proof-credibility-assessment-eu-asylum-systems. factor in their asylum claims. 30 The prevalence of SGBV in the lives of women migrants and refugees is thus increasingly well documented.
Fleeing SGBV is also the reason why many women migrate in the first place. Despite this, and despite international and regional frameworks that establish SGBV as a violation of human rights that can constitute persecution, asylum claims involving experiences of SGBV are frequently deemed 'insufficient' to warrant protection. 31 Women asylum-seekers are commonly viewed as victims of crime rather than persecution, due to the typically 'private' and 'non-state' nature of the violence. Sexual violence is particularly ambiguous territory since it occurs everywhere; it is thus challenging to decipher whether the home state has the will and capacity to offer protection from it. In the UK asylum system, several studies drawing on women's accounts 32 and interviews with key informants 33 have found that women applicants are met with 'cultures of disbelief' linked to racialised gender stereotypes. Drawing on a case study, Grøvdal and Bjørnholt similarly demonstrate reluctance to protect trafficking victims in the Norwegian asylum system. 34 These challenges may be further exacerbated by the criteria used in the credibility assessment, which expects details, specificity, and consistency. Disclosures of SGBV are complex because they are associated with trauma and shame and may lead to disruptions within families and communities, including the shunning of the victim. Experiences of rape and sexual violence can also result in diagnosable post-traumatic stress disorder, which critically impacts a survivor's ability to disclose the experience(s). 35 Most survivors of SGBV do not report it, in any country or context. 36 They usually do their utmost to avoid thinking of those experiences; even with professional help, it may take years or decades before they can tell their stories 37 and when they first do so, the events are typically excused or minimised. 38 Suppressed aspects can be triggered by reminders, e.g. when applicants are asked to disclose their experiences in the asylum interview. They may then re-experience the traumatic event(s) and, as they are processing their trauma with each retelling of the story and gathering images and sensations from the experience, the narrative might be fragmented, disrupted, and inconsistent. 39  narrative of the event signals recovery, and a full and detailed account may never be available in the survivor's memory. 40 The credibility criteria may thus be at odds with how traumatic experiences are usually disclosed, 41 particularly cases of SGBV. Nevertheless, few studies have explored the association between asylum credibility assessment and disclosures of SGBV. Baillot et al. are unique in having explored the link between asylum credibility assessment and disclosures of sexual violence in the UK asylum system. 42 They find that significant structural hurdles face women applicants whose asylum claims contain rape allegations, such as gendered stereotypes, the impact of trauma on their narratives, and an unwillingness among caseworkers to engage with the allegations.
Whereas Baillot et al. focused on the judicial side of the asylum proceedings, the focus here is on the credibility assessments conducted by caseworkers. This focus allows us to home in on the daily work and procedures of these street-level bureaucrats. The following section suggests that caseworkers may rely on mechanisms that allow for simpler decision-making. This might negatively impact women applicants' ability to disclose their experiences of SGBV and, when they do, increase the risk of such disclosures being disregarded.

Theoretical framework: street-level bureaucrats
Street-level bureaucrats are public officials who interact directly with clients 43 when implementing public policies. Their work usually involves discretionary power; 44 they must adhere to rules and directives from above, while assessing human dimensions too complex to reduce to a procedural format. 45 Bureaucrats must treat all clients in common circumstances alike, but also respond to their unique situations. Arguably, it is challenging to make such complex decisions under what are often large caseloads and with limited resources. 46 To cope with this reality, Lipsky argues that street-level bureaucrats may deny discretion by strictly adhering to the rules and emotionally detaching themselves from clients. 47 They establish routines and adopt mechanisms that allow them to psychologically simplify their environments and clients, thus facilitating a mass processing of cases. Two common mechanisms are categorisation and standardisation. These are used in the name of fairness and equity, yet street-level bureaucrats are often confronted with the apparent unfairness of treating people alike, because individuals are more than their assigned category. 48 The very transformation of a unique individual into a client is a social construction. People-processing agencies develop a set of categories to sort people, and then map clients according to their (dis)qualifying characteristics. 49 When categorising clients, bureaucrats tend to treat them as bundles of bureaucratically relevant attributes rather than as human beings, which increases the likelihood of missing important dimensions of the presented problem. 50 Lipsky also shows that bureaucrats can absorb information and standardise clients according to the preconceptions and expectations for each category; these then become institutionalised, as each case sets standards and establishes procedures for future treatment of similar cases. 51 Their routines and mechanisms thus influence decisions and become actual policies. Interviews with clients are typically structured to expedite information-gathering and are standardised with the assumption that a new client will fit the contours of previous clients in that category. 52 Clients may attempt to tell their full stories in their own ways, but these efforts are typically shut down by the bureaucrats. This standardisation can impede bureaucrats from discovering unique circumstances, which would require more flexible responses. Thus, having restricted the kind of information they receive, street-level bureaucrats often find confirmation that clients tend to fall into well-defined categories and that the standardisation of their claims is fair and efficient, thus influencing their notion of whether clients are deserving of help and compassion. 53 Since Lipsky introduced the concept of street-level bureaucracy in 1980, and it has been thoroughly studied, 54 mostly in contexts where street-level bureaucrats are responsible for distributing public goods and granting access to government programmes. Coping mechanisms such as limiting client demands or creaming/cherry-picking have been emphasised. 55 Previous studies have also found coping mechanisms to be utilised in asylum bureaucracies. 56 Not all coping mechanisms will be equally relevant in every context, however. In an asylum bureaucracy, the rather limited previous research presented below indicates that caseworkers are more inclined to adopt mechanisms of categorisation and standardisation.
Studies of street-level bureaucracy theory in credibility assessment processes Liodden lays out a comprehensive picture of the complexities of asylum decision-making and the ways in which caseworkers in the Norwegian Directorate of Immigration ('Utlendingsdirektoratet', UDI) manage their discretionary power. 57 She finds that they strive to treat each applicant as a distinct person with her own story, yet frequently utilise categorisation and standardisation. Asylum-seekers are categorised with others from the same country/social group, for example, as caseworkers adhere to the established 'practice', a system that provides guidelines and notes based on interpretations of the law, political instructions, available country information, and the most common asylum claims among applicants from each country, as well as decisions in similar cases. Based on the 'practice' system, caseworkers can approach cases with a hypothesis of truthfulness or disbelief, with the latter seemingly alleviating some of the distress of making difficult decisions. 58 Bollingmo et al. equally find that some caseworkers at the UDI develop a hypothesis about credibility beforehand and use the interview to confirm their assumptions. 59 These findings are in line with those made by Jubany and Dahlvik, 60 who observe that asylum caseworkers in Spain, the UK, and Austria predominantly rely on their 'intuition' or 'gut feeling' in their credibility assessments. During interviews, these caseworkers use tactics to spot or trigger inconsistencies in applicants' stories. Dallara and Lacchei confirm the use of both organisational and individual coping mechanisms among asylum judges in Italy. 61 They emphasise the validity of applying the street-level bureaucracy framework to asylum systems.

Theoretical expectations
Based on Lipsky's seminal work, we will explore whether and how asylum caseworkers apply the mechanisms of categorisation and standardisation in credibility assessments. Knowledge is lacking on the daily decision-making processes of this group, although the challenges and mechanisms identified by Lipsky might be particularly salient in a context where bureaucrats must make decisions that will determine applicants' futures, based on the stories they tell. We expect that this context may adversely affect whether and how disclosures of SGBV are enabled and, if disclosed, the weight they are ascribed. The overall aim of this study is therefore to investigate whether women's claims are considered in their entirety and their human right to protection is upheld in practice.

Methods and Data
The Norwegian case We have chosen a deep dive into the Norwegian case in order to acquire a comprehensive understanding of one aspect of the global migration regimethe asylum systemat the national level and explore how international asylum laws and procedures are localised and 'at work' in daily life. 62 While many refugees come to Norway through resettlement 58 Liodden 2016 (n 12). 59  schemes, the focus of this study is asylum-seekers. Between 2015 and 2021, 44,515 asylum-seekers came to Norway to seek protection as refugees. 63 In 2020, 52% were women and children. The applicants must be on Norwegian ground in order to apply for asylum. They must do so by registering and presenting their claims with the Immigration Police (PU, 'Politiets Utlendingsenhet'). They are offered information from the Norwegian Organisation for Asylum Seekers (NOAS) before being moved to a transit centre pending further process. For families applying together, the adults are typically interviewed separately.
The Norwegian Directorate of Immigration (UDI) is responsible for making asylum decisions in compliance with policies and legislation on international, European, and national levels. Regarding whom should be recognised as a refugee, Norway adheres to the Refugee Convention and its 1967 Protocol and utilises the UNHCR Handbooks as guiding documents. 64 If an applicant does not meet the criteria of the Refugee Convention, s/he can still be considered a refugee if there is a 'real risk of death penalty, torture or other inhumane and degrading treatment or punishment upon return to the country of origin', or considered for a permit on 'strong humanitarian grounds' (Norwegian Immigration Act, §28b, §38); the principle of non-refoulement is essential (ibid., §73).
Caseworkers conduct the asylum interview, which follows the 'free narrative' methodology, widely recognised as best practice. The applicant can speak freely, and the caseworker will follow up, ask for clarifications, and take verbatim notes. 65 The interview is also recorded. The caseworker makes the initial asylum decision, but the interview report and the decision are subsequently reviewed by a second caseworker. The UDI aims to ensure that caseworkers consider evidence and credibility in a harmonised manner. 66 The UDI adheres to the credibility criteria offered by UNHCR: 67 sufficiency of detail and specificity; internal consistency; consistency with information provided by family members and/or other witnesses; consistency with available external information; and plausibility. Contradictions and ambiguities must be justified in a 'satisfactory manner' and late disclosures may weaken overall credibility, particularly after the initial claim has been rejected. 68 It is at the discretion of the caseworkers, who typically have educational backgrounds within political science or law, to determine whether an account is credible and 'reasonably likely'. 69 Caseworkers at the UDI, particularly those who have joined in recent years, complete a standardised training period. They are informed of the guidelines for interviews and credibility assessments. While the trainings and guidelines mention that applicants may have experienced violence they are reluctant to talk about and that trauma may impact their ability to tell their stories, there is no focused training for staff on how to handle cases of SGBV.
Should the claim be rejected, the applicant has the right to free legal counselling and to submit a complaint to the UDI. 70 If the UDI upholds their decision, the claimant can appeal the case to the Norwegian Immigration Appeals Board ('Utlendingsnemda', UNE), a court-like body whose board members have law backgrounds and competencies as judges. 71 The decisions at the UNE serve as precedents for both bodies. 72 In 2019, the UNE overturned 17% of asylum appeals. 73 The Norwegian asylum system can be considered a street-level bureaucracy. The caseworkers/bureaucrats are free to make their own decisions, within the confinements of the legal rules, and they must make a definitive decision in often uncertain circumstances. 74 Applicants cannot avoid encounters with a caseworker if they wish to obtain protection, and power lies entirely with the caseworker.
Although this case study is limited to Norway, its findings could also prove useful for other asylum contexts. European countries strive for a common asylum system where applicants should be treated fairly and equally. 75 Every host country will receive women applicants and it is likely that many of them will have experienced SGBV. An increased understanding of this phenomenon in one host country could thus be beneficial for other countries. Additionally, an underlying expectation of this paper is that street-level bureaucrats will experience similar challenges and utilise comparable coping mechanisms, regardless of the context.
Norway is also unique in some respects, however. It ranks among the top countries regarding gender equality, which makes it an interesting case given the study's focus on women applicants. 76 Norway also adheres to several 'best practices' identified in the literature on gender and asylum, such as the free narrative approach in asylum interviews. Furthermore, the applicant can solicit free legal advice and request a female interpreter or interviewer. 77 Norway adheres to the international laws and conventions on SGBV and gender-related persecution, and 'vulnerable women' are a stated priority in Norway's immigration policy. 78 If we can identify barriers to disclosures of SGBV and subsequent protection on such grounds here, it must be expected that there are significant hurdles in other asylum contexts.

Data
The main empirical material comprises qualitative semi-structured interviews with 18 key asylum actors and 85 public case summaries from UNE's 'Praksisbase'. 79 We will describe this data separately below.

Interviews
The interviewees are UDI caseworkers, UNE board leaders, immigration lawyers with work experience as UDI caseworkers, an NGO employee with critical knowledge of the Norwegian asylum system, and health professionals working with traumatised refugees. Five participants are men and 13 are women, with experience ranging from three to 20 years. The sampling was done purposefully to capture a holistic view of the asylum system, while adhering to the primary focus of the study by predominantly interviewing UNE and UDI caseworkers. The interviews were conducted by the first author in 2020.
The participants received information about the study beforehand, as well as in the beginning of the interview. They were also made aware that their participation was voluntary and that their consent could be withdrawn at any stage. The questions focused on the subjective process and reflections of the caseworkers pre-and during an asylum interview, and the subsequent decision-making: potential hypotheses established beforehand, what makes a story credible, the impact of displays of emotion, how the vulnerability of women applicants is assessed, the relevance of disclosures of SGBV, and caseworkers' discretion and decision-making.
The interviews were recorded and manually transcribed and coded. The analytical process was abductive; we gave room for the data to speak and unexpected elements to be discovered. 80 Our expectation, however, was that the coping mechanisms suggested by Lipsky, 81 including categorisation and standardisation, would be discernible in the caseworkers' reflections on their work. As these mechanisms are often unintentional, we did not explicitly ask about them during the interview, but rather looked for cues during the analysis that could be associated with such mechanisms.

Database
When a case and subsequent complaint is rejected by the UDI, the applicant can file a complaint with the UNE. The Praksisbase database provides short summaries of the UNE justifications for either overturning or sustaining the UDI decision. For this study, 85 of these case summaries were analysed (40 overturns and 45 rejections). These were all cases in 2015-2020 that mentioned gender-related persecution. Additional search terms were 'sexual violence', 'trafficking', 'rape', 'domestic violence', 'FGM', and 'forced marriage'. We also reviewed the UDI guidelines pertaining to credibility assessment, decision-making, and gender-related persecution. We used the database summaries as a complement to the interviews, previous literature, and the UDI guidelines to see how the identified coping mechanisms may materialise in practice.

Data limitations
We could not access the UDI's anonymised cases, due to ethical concerns regarding confidentiality and anonymisation. It is thus important to consider that the findings related to the UNE's Praksisbase are solely based on cases that were initially rejected by the UDI and then brought forth to the UNE. We have not reviewed UDI cases where the applicant was found credible and offered asylum, and although our analysis relies predominantly on the interviews, this caveat should nonetheless be considered when evaluating the study's findings and generalisations.
An additional limitation is that this research does not include interviews with applicants. This is partly due to difficulties in obtaining access, but it is predominantly a deliberate ethical decision. Considering the often-traumatising nature of disclosures of SGBV, we wished to minimise the need for further retelling. A potential criticism of this study is, however, that it perpetuates further silencing of SGBV survivors. Future research should aspire to give voice to these applicants on their own terms.
An important caveat to this study is that male migrants and refugees are also subject to sexual violence. This paper exclusively addresses women applicants because women migrants and refugees are disproportionately affected by sexual violence. 82 Given the highly gendered nature of the phenomenon, the victimisation of men and women should receive separate attention. We expect, however, that many of the findings of this paper also apply to male applicants who have experienced sexual violence, particularly pertaining to the difficulties of disclosures and associated shame and trauma.

Analysis
In this section, the findings are presented and analysed using Lipsky's theory of streetlevel bureaucrats. The overarching question is whether there are any barriers to disclosures of experienced SGBV and, subsequently, to obtaining protection on such grounds, in the Norwegian asylum system. The analysis focuses on the coping mechanisms of standardisation and categorisation.

Standardisation of disclosures
All caseworkers emphasised that their primary function was to assess the grounds for protection. As they usually knew what they had to establish during the interview to accomplish this, certain aspects would be more relevant than others. The caseworkers all referenced the significant prevalence of SGBV in women applicants' lives, but many added that there would not be sufficient grounds for asylum unless there was a risk of future persecution. Thus, dwelling on it was pointless: This is an enormous dilemma regarding women applicants. They have often been subjected to horrible things. Sexual violence in conflicts or family situation. However, if there is no risk of future persecution upon return, it is pointless discussing these experiences. It is horrible to say […] it will be entirely irrelevant for the outcome of the asylum decision. (UNE board leader 1) A lawyer and former UDI caseworker added that the Norwegian asylum system should consider this as pertaining to vulnerability upon return but emphasised that this was not the current procedure. Inquiring into such experiences would thus just add to the trauma: if a woman must sit and talk to me about having been raped on a boat on the Mediterranean, when it will not matter to the final decision, that is merely tormenting that woman. (Former UDI caseworker 2) 82 UN Women 2021 (n 7).
There was also a consensus among the caseworkers that the experience of SGBV, particularly sexual violence, was a difficult topic to approach and many appeared to prefer to leave it out altogether. Several caseworkers mentioned that they were not psychologists (e.g. UDI caseworker 1, 4, and 6). One said that 'you should not ask questions to an excessive degree, because you might stir up something that you are not qualified to handle' (UDI caseworker 4). Inquiring into experiences of SGBV, she explained, could potentially 'open a Pandora's Box and trigger reactions that we neither can, nor should, try to handle'.
Another reason for not inquiring into SGBV experiences was to keep suffering at a distance 'in order to sleep at night' (Former UDI caseworker 2). As Lipsky theorised, 83 caseworkers used emotional detachment as a coping mechanism in their work: When you hear the stories too many times, it does something to you. After a while, I could not bear hearing them unless I had to. So I did not create space for them to tell these stories. In hindsight, that was probably bad. Maybe they felt a need to tell; finally, someone would listen. (Former UDI caseworker 1) The applicant might also be unwilling to share her experiences of SGBV due to shame or the risk of being shunned by her community and family. According to one UNE board leader (2), experiences of sexual assault were often not disclosed because the applicant was 'terrified that the husband would hear and kick her out'. Such disclosures require trusting and enabling environments, which take time to build, particularly for women in an asylum context who may be fearful of authorities. The fact that caseworkers should only spend one day on the interview is therefore not likely to facilitate disclosures of SGBV. Caseworkers can complete a second or third interview, but the degree to which this is done seems to vary between caseworkers.
An important aspect to consider here is that Norwegian caseworkers should not just assess whether an applicant meets the grounds for protection, but also if internal relocation options are feasible, and whether the applicant should be offered a residence permit on humanitarian grounds (Norwegian Immigration Act, §38). An Amnesty International senior advisor described SGBV as 'grave violations of human rights' and explained that caseworkers' reluctance to probe into these experiences might result in a woman's claim not surfacing in its entirety. She insisted that they should facilitate disclosures of such experiences to assess whether the applicant should be given a residence permit on humanitarian grounds or if she needs professional help in Norway before she is returned.
Instead, a recurring critique in the literature and in some of the interviews is that women's voices are barely heard at all. Women who arrive with families often do not receive dedicated attention and those who arrive alone are frequently judged based on their proximity to men. 84 One lawyer and former UDI caseworker explained: The asylum system is quite patriarchal […] the man is automatically considered the main actor. He's the one awarded a lawyer, and then the wife and children are add-ons. I think this results in us focusing far more on the men and not being as attentive to what happens to women, or their experienced persecution. The families themselves often come from relatively patriarchal societal systems so often they too focus on the man's need for protection. He's the one who has been out acting in society and been politically active. (Former UDI caseworker 2) The aspects discussed thus far indicate a standardisation of the disclosures of asylum claims. The caseworkers said they had to focus on the aspects of the applicant's story that were important for the asylum decision and that they were not qualified to handle disclosures of SGBV. They appeared to merely probe the standardised aspects of the claim, rather than the full experience of the applicant. This standardisation may result in disclosures of SGBV not being encouraged or enabled, despite the many reasons why this information might be valuable, even if it does not fit the criteria for protection.

Standardisation of credibility
The previous section showed that the caseworkers interviewed avoided discussing experiences of SGBV. Where such experiences are disclosed, however, there is a risk that the way in which the disclosure occurs is at odds with the pre-established criteria that caseworkers look to when assessing the credibility of any asylum claim. The credibility criteria state that the applicant's narrative should be coherent, consistent, and detailed. Any anomalies should be justified. The UDI guidelines (2020) acknowledge that ability to meet these criteria will differ depending on individual circumstances, but it is at the discretion of the caseworker to determine how much deviation is acceptable.
In 21 of the 45 reviewed case summaries of rejections in the UNE's Praksisbase, 85 there were references to explanations lacking credibility because they were 'vague', 'lacking in detail', 'shallow', 'imprecise', 'inconsistent', or 'contradictory'. Former UDI employees claimed that many caseworkers in such situations rely on 'crutches' or 'pegs' (former UDI caseworker 2) that can 'justify rejections' (former UDI caseworker 1): I think the caseworkers like to have their 'practice', because being in doubt is so difficult. So having this 'crutch' or 'peg' to rely on and say that 'no, we can't attribute weight to this part of the story, we can't believe it, a book tells me that I can't, so then I don't have to believe it, then I don't have to live with doubt. The doubt that might lead to me being tricked'. No one likes being tricked. Nor do they like walking around thinking they have rejected someone who should have been offered protection […] and this is what I believe to be one of the major issues in the Norwegian asylum system, that caseworkers to such a large extent rely on these to avoid utilising discretion, because doing so is difficult and uncomfortable. (Former UDI caseworker 2) The 85 UNE case summaries reviewed contained only one reference to trauma (N1612791205), despite many references to vague and inconsistent explanations. In this case, the claimant alleged that she had been kidnapped, detained, and abused by the Taliban. The summary stated that her explanation was 'illogical', 'fumbling', and 'unclear' and that had significant inconsistencies with previous explanations. Some board members believed that the reason could be the 'complainant being traumatised and having difficult memories due to her experiences', but the majority did not find her story credible, and she was thus not offered protection.
Caseworkers appear more inclined to rely on the credibility criteria when a case is less 'clear cut'. The UNE legal advisor explained that 'if what the person is telling has never happened to anyone else, it will most likely weaken the credibility'. Late disclosures also typically weaken credibility. Most caseworkers acknowledged that there may be ample reasons for late disclosures. Several caseworkers mentioned, for example, that most migrants arrive in Norway without knowing the requirements for receiving asylum or that their experiences of violence may count as persecution. 'To them, it is so normalised' (e.g. UNE board leader 1 and 2; UDI caseworker 4). The information session with NOAS covers these topics, but this only occurs after applicants have registered with the PU and they may thus not think to mention certain experiences of violence in the initial claim. Inconsistencies between the police statement and the asylum interview may weaken credibility and any inconsistencies must be explained: why is it surfacing now? Can you give an adequate explanation as to why it's coming late? The further into the process this emergesespecially after a rejectionif you then disclose that 'yes, I've been subjected to sexual violence', well, it takes a lot to believe that, unless one can give a reasonable explanation. Then comes the challenge of deciding whether the explanation is truthful. (UNE board leader 1) An asylum lawyer and former UDI caseworker (2) expressed frustration with what he perceived as a mantra in both UDI and UNE decisions about late disclosuresthat they are less credible. According to him, this mantra reveals a 'poor understanding of what shame does to a human' and how it may make someone reluctant to speak about traumatising experiences. He also criticised the naivety of Norwegian authorities, claiming that they assume of applicants that 'Surely, they must trust those they are seeking protection with, so they must be able to disclose everything early on, preferably in the first conversation with the police, but at least in the asylum interview'. Quite the contrary, he said, it is only logical that these disclosures surface after a rejection: they had believed that the story they had told would be enough. That one part of the story, or maybe the fabricated story, the one they fabricated because they felt unable to talk about the thing that felt too dangerous or too shameful. Then they are told that it was not enough […] they might think that 'OK, I guess I need to tell them about all the painful stuff that I'm so ashamed of, the things I never wanted to tell anyone'. They then bring forward this complaint after the rejection and then they lose even more credibility. Expanding one's story in connection to a rejection and complaint is the death of credibility. (Former UDI case worker 2) A case summary from Praksisbase (N169059114) showed that what is considered a valid reason for late disclosures and inconsistencies varies. In one case, the applicant feared forced marriage upon return, but she had not disclosed this in the police registration. She explained to the UDI that she had not done so because the police told her that she should only answer the questions asked. Additionally, she had been recovering from an abortion and had spent the entire night sitting on a chair in the PU's offices, as the only woman, and had therefore not slept. The UDI deemed her explanation for the inconsistencies implausible, but the UNE disagreed and overturned the decision.
The caseworkers interviewed all stated that applicants' abilities to meet the credibility criteria differ, but that decisions nonetheless rely on those criteria. As Norway operates with a free evidence assessment, it is up to the caseworker to decide how much 'evidence' is required for a claim to be considered credible. The threshold for what is considered credible thus differs, as the credibility criteria do not specify how internally consistent or how plausible the applicant's account must be. This appears to prove somewhat of a paradox, which can be related to Lipsky's theory. 86 Caseworkers attempt to limit discretion by relying on the credibility criteria; they strive towards a standardisation of their decisions, but as there is no consistent rule regarding how much deviance or adherence is acceptable, the decision still relies on discretion.
To summarise, examples of standardisation can be identified in various stages of the decision-making process. In the asylum interview, there is a standardised view of which aspects of a case are relevant to determine the grounds for protection. SGBV typically falls outside of these parameters, and the caseworkers thus do not enable such disclosures. Consequently, caseworkers also shield themselves from difficult stories and situations that they do not feel qualified to handle. Applicants, for their part, may also be unwilling or unable to talk about these events. There are thus several factors inhibiting women applicants from giving a full account of their experiences, especially in the initial stages of the process. This becomes a problem in the decision-making stage, or when the decision is contested, as the standardised response to late disclosures and inconsistencies is scepticism.

Categorisation of countries and ideal victims
We identified two kinds of categorisations that are closely interlinked: one of countries and one of ideal victims. Categories of countries from which applicants come are critical in the Norwegian asylum context, as all cases are treated according to 'practice' and the standard for each country, which leads to a standardised response to all claims in that category. Some origin countries lead to a nearly 99% approval rate for applicants, while others have a less decisive impact on applications. Credibility typically becomes essential for applicants coming from the latter countries; there is more room for discretion and thus decision-making is more difficult.
The findings suggest that hypotheses are developed for applicants from particular countries and that their claims are standardised accordingly. An example of such a hypothesis is that imposed on applicants from Afghanistan claiming to be single women without a male network. One UDI caseworker explained that the journey to Norway from Afghanistan is so difficult and expensive that it is unlikely that a woman without a male network could fund and organise it. According to her, caseworkers had thus developed the hypothesis that either (a) a male network had helped the woman and thus cared enough about her to protect her back home, or (b) that she was 'resourceful enough' to make it to Norway on her own. Hence, she was not a genuinely 'vulnerable woman' who really needed asylum. The same hypothesis was applied to all women applicants from Afghanistan: In these cases, there are many things that indicate that something is off. Even though it may be difficult to say in individual cases that this aspect is implausible, the overall picture of all these cases shows that what they are saying is not correct (UDI caseworker 7).
Similarly, a former UDI caseworker revealed in the newspaper Dagbladet in 2019 that UDI caseworkers had received instructions that they should assume that all women arriving from Afghanistan who claimed to be 'single without a male network' had the necessary network back home, 'regardless of what they themselves said'. 87 The categorisation of women applicants based on their country of origin is thus closely intertwined with the extent to which they are categorised as 'vulnerable' and 'ideal victims'. 88 'Vulnerable women' are a priority in the Norwegian immigration policy. Being categorized as such increases a woman's chances of receiving protection; not fitting into the stereotype of a passive refugee victim may have the opposite effect. 89 Jobe found in her study of women seeking asylum from trafficking that only those placed in the category of 'ideal trafficking victims'innocent and 'respectable' women who had been clearly coerced and victimized received protection. 90 Women applicants were frequently judged based on their past sexual experiences, how much agency they seemed to have displayed in the situation, and the perceived willingness with which they may have entered the sex industry.
Labels associated with being 'vulnerable' and an 'ideal victim' can evoke expectations of certain emotions and demeanours from the victim. These expectations have been observed in studies of rape allegations in the criminal justice system. 91 One former UDI caseworker also confirmed this phenomenon in Norwegian criminal courts and compared this to the asylum system: and still, there [in criminal courts], they have trained prosecutors. In asylum interviews, you're just throwing regular people out there, with their own convictions that aren't necessarily based on facts, who will decide whether they think someone is telling the truth: 'She has to cry; she has to be unable to speak because she's so upset'.
This answer to whether displays of emotion could affect the credibility assessments further exemplified how a categorisation of applicants can influence caseworkers' expectations: This is an enormous weakness in the asylum institute. I try to be very aware of this […] these things are culturally contingent. It bounces right off when an Iranian woman is sobbing and about to faint. They're raised to behave this way and it's expected. (UNE board leader 1) Similarly, a UDI caseworker (1) stated that: when I first began, I was very surprised when I interviewed women from Somalia who said straight out: 'I was raped' and then just carried on with their stories. This was just a minor detail in the story because it was so common and normalised. If one isn't attentive to it, it's easy to think that they're too nonchalant about it and that it's not that serious […] This comes with experience, the first time one might think that the person seems indifferent and that might seem very strange […].
The quote above reveals expectations of emotional displays and that the credibility assessment may be impacted by whether the applicants meet these expectations. Some caseworkers underscored that they did not allow displays of emotion to affect their assessments, but they believed that less experienced caseworkers did (e.g. UDI caseworker 1; UNE board leader 2). The UDI guidelines (2020) emphasise that displays of emotion should not be used as indicators of credibility. Yet a UDI caseworker (5) doubted that these were always followed: you have others who can sob through an entire asylum interview, and that's a lot, right? Where you feel it's an act. Others can talk about gruesome events and be entirely flat. Even if they don't display any emotions, you can almost think it's credible for that reason. I therefore don't believe it's right to say that this isn't a factor.
There are thus indications that some caseworkers, perhaps unconsciously, have expectations regarding emotional displays, but that what constitutes the 'correct' level appears to be subjective and dependent on which category the applicant belongs to. As such, displays of emotion by a woman applicant when disclosing experiences of SGBV may affect the perception of her credibility and the subsequent decision. These preconceptions and hypotheses appear to be associated with categorisations of applicants based on country of origin.

Conclusions
The aim of this paper has been to explore whether and how disclosures of SGBV are encouraged and considered in the Norwegian asylum process. Previous research has found that many women come up against significant barriers in attempting to obtain protection under the Refugee Convention, but this has rarely been explored in connection with the credibility assessment and disclosures of SGBV. The international and regional human rights frameworks firmly establish SGBV as a violation of human rights law that may constitute gender-related persecution and thus give grounds for protection. Nevertheless, our findings suggest that this may not materialise in practice.
By analysing the personal processes and reflections of the caseworkers behind credibility assessments and decision-making, this paper shows how two mechanisms that caseworkers use to simplify decision-makingcategorisation and standardisationmay directly and indirectly impact women applicants' ability to tell their full stories and subsequently to obtain protection on the grounds of SGBV. Our study thus confirms the fruitfulness of the street-level bureaucracy framework in exploring how asylum policies are interpreted, applied, and reinforced in practice.
There are many good practices to be identified in the Norwegian case. The UDI interview follows best practice, with only the caseworker and an interpreter present. The caseworker is instructed to put the applicant at ease by being friendly, to provide safe and predictable structures for the interview, and to allow the applicant frequent breaks and the space to tell his/her story uninterrupted. These things are critical for traumatised applicants' well-being and for disclosures of difficult experiences to take place. It is also our distinct impression that the UDI and the UNE aim to take women's particular needs into consideration, especially by ensuring the use of female interpreters and interviewers. Furthermore, the caseworkers themselves appear to go to great lengths to establish the best human connection possible with the applicants within the given time frame.
We find, however, that disclosures of SGBV are not currently being facilitated by some caseworkers, as they perceive such experiences as falling outside the parameters of the standard grounds for protection. They thus refrain from probing this sensitive topic for fear of causing unnecessary pain and uneasiness to the asylum-seekers and themselves. This result complements similar findings from the UK asylum system. 92 The standardisation of how credible disclosures should be made also appears to be at odds with the way SGBV experiences tend to be revealed. Late disclosures, inconsistencies, and a lack of detail are known to be entirely common in the narratives of SGBV survivors. Our interviews testify to the difficulty women asylum-seekers, just like other survivors of SGBV, may face when talking about experiences which can be closely connected to trauma and shame. Applicants may thus refrain from disclosing such experiences until they absolutely must, e.g. after a negative asylum decision, when the standardised notion of credible disclosures may be used by caseworkers to justify another rejection. Caseworkers' attempts to limit their discretion by relying on credibility criteria is paradoxical, however. Because it is not predetermined how consistent a narrative must be with the criteria, the decision still relies on discretion and there can be great variations in outcomes across caseworkers.
While we recognise that disclosures of SGBV can be difficult and re-traumatising for an applicant, and that the choice of whether to disclose or not is naturally hers, we identify salient reasons for why caseworkers should attempt to enable such disclosures. While Norway has committed to ensuring the protection of 'vulnerable women', the full picture of vulnerability may not surface: the limited time spent on asylum interviews is not sufficient to enable disclosures of traumatic experiences, particularly if the caseworker is reluctant to hear them. Women who could be eligible for protection/residence permits may thus be rejected and returned. Additionally, women who are offered protection might not receive the help that is critically needed after experiencing SGBV.
Furthermore, choices about who to categorise as a 'vulnerable woman' are closely related to notions of 'the ideal victim' and a woman's nationality/ethnicity. Like previous researchers, 93 we find that caseworkers' decisions may be influenced by the extent to which a woman is perceived to be innocent, passive, and lacking in agency. Caseworkers base their perceptions of these characteristics on a combination of the woman's nationality/ethnicity and the degree of emotion she displays while recounting her experiences of SGBV. Despite these attempts to categorise asylum-seekers, the caseworkers interviewed point out that asylum decisions are likely to vary between caseworkers. This study thus underlines the critical role of caseworkers in the asylum processes.
There are strong reasons for asylum institutions to invest in further training of caseworkers on trauma and SGBV, particularly sexual violence, and how trauma may adversely impact an applicant's ability to present her narrative in accordance with credibility criteria. This would be beneficial for both male and female survivors of sexual violence and other human rights abuses. SGBV is prevalent in the lives of many who migrate. The Norwegian authorities should strive to acquire critical insight into the scope and consequences of their experiences, to better offer help to those who are currently in Norway, whether they are there to stay or not.
To conclude, this paper contributes to the growing body of literature that encourages conversations which recognise the importance of enabling disclosures of SGBV. By focusing on credibility assessments and SGBV in the asylum context, we hope to prompt a recognition of the current barriers blocking women from obtaining protection, or residence permits, on such grounds in Norway and beyond.

Additional Information
The article is part of the project 'Violence against women migrants and refugees: Analysing causes and effective policy response', funded by Horizon 2020 and the Research Council of Norway as part of the GENDER-NET Plus consortium.
The project has been recommended by the Norwegian Centre for Research Data, NSD, Project Number 786140.