Locating the Concept of Vulnerability in Canada’s Refugee Policies at Home and Abroad

How is the concept of “vulnerability” employed in Canadian immigration law? This article presents findings from research conducted as part of the VULNER project (2019–23). We analyze how vulnerability is operationalized in Canada’s inland refugee (or asylum) determination procedures compared to its overseas resettlement program by first discussing some general principles, followed by an examination of the treatment of women and LGBTQI+ individuals seeking protection. We show that legal-bureaucratic logics have constructed two very heterogeneous worlds of vulnerability that may reproduce and exacerbate vulnerabilities rather than create a more inclusive, equitable protection regime in law, policy, and practice.


Introduction
What does it mean to be "vulnerable" in Canadian immigration law? In other words, how is the concept of "vulnerability" employed in law and policy making? This article presents findings from research conducted as part of the international VULNER project . More specifically, we analyze how vulnerability is operationalized in Canada's domestic (inland) refugee (or asylum) determination procedures compared to its overseas resettlement program. Such a comparison is novel as these two components are rarely considered together. One of the reasons is that Canada's well known refugee resettlement program is a humanitarian admission scheme, administered by the Ministry of Immigration, Refugees and Citizenship (IRCC), that operates largely outside of the rule of law (Labman 2019), while its inland refugee (or asylum) determination procedures (RSD) represent Canada's legal commitment to refugee protection as a signatory state under the 1951 Geneva Convention for Refugees (the Convention) and the 1967 Protocol. Inland claims are adjudicated by the Immigration and Refugee Board (IRB), Canada's largest administrative, quasi-judicial tribunal, which was created in 1989 and continues to operate independently of IRCC. A second reason why not many academic works consider both aspects of Canada's refugee policy, is that much less is known empirically about Canada's resettlement procedures compared to its inland refugee determinations, although this is beginning to change (Garnier et al. 2018, p. 3). A third reason is that contemporary resettlement decisions are less transparent and more discretionary, less frequently reviewed in the courts, and as a result, overall opaquer; and therefore, more challenging to analyze.
Nevertheless, we believe that such a holistic analysis of the concept of vulnerability as it is operationalized in both components of Canada's refugee policy is timely, given that at the global level, shared commitments to address migrant vulnerability in state policies and practices have been steadily increasing, for instance in the 2018 Global Compact for Safe Orderly and Regular Migration (Atak et al. 2018). Finally, given Canada's historic leadership in the global refugee regime (Milner 2021), there are good reasons to expect that Canadian laws, policies, and practices may hold valuable lessons for scholars and policy makers 1 All VULNER project partners also interviewed immigration lawyers, government decision makers, frontline practitioners, and migrants themselves. At the time of writing, the Canadian team had not yet concluded its interviews. 2 Our discussion of recent case law is not meant to be exhaustive or conclusive. It is merely used for illustrative purposes and draws predominantly from the work of Anna Purkey in this special issue.
violations of their rights" (International Organization for Migration 2017). Yet which characteristics or situations warrant what type of protection status in law, or what type of procedural accommodations, and who or what is triggering the exposure to these risks is far from clear, as "vulnerability" remains conceptually muddled in both law and policy. On the one hand, vulnerability is typically understood as being caused by inherent, embodied characteristics at the individual level, such as someone's age or health. On the other hand, it can also be evoked by certain situational characteristics, which are often experienced by entire groups of people, such as their language or ethnic origin. Furthermore, whether someone is rendered vulnerable and ends up experiencing harm or is only potentially at risk (for instance, of being trafficked), can lead to very different (and at times contradictory) policy responses. At the same time, the concept of vulnerability is also employed normatively (Brown et al. 2017, p. 498), identifying only certain people worthy of protection, often those who do not (or cannot) exercise any agency (Ataç et al. 2016). This can lead to a vulnerability 'contest,' as some vulnerabilities become normalized or valued more than others (Howden and Kodalak 2018). Further, state policies may render individuals equally vulnerable, either in their home country or during their journey to safety, blurring the boundary between citizen and migrant and possibly leading to a humanitarian 'fatigue' (Gallant et al. 2019). Given the global trend towards the securitization (Crépeau and Atak 2013) and criminalization of migration (Atak and Simeon 2018), together with the extraterritorialization of borders (Côté-Boucher et al. 2014), a commitment to addressing and reducing vulnerabilities is even more urgently needed.

From Vulnerability to Precariousness: The Role of the State
Martha Fineman's (2008) seminal work on 'the vulnerable subject' is regularly referenced in current writings on vulnerability and the law (e.g., Baumgärtel 2020, p. 14). While recent contributions have advanced Fineman's critique on human subjectivity, her understanding of the role of the state has been less frequently analyzed. We argue that states and state actors need to do more than become "more responsive" when faced with vulnerabilities, (Fineman 2008, p. 13), in part because they, as "precariousness" scholars have shown, are implicated in actively constructing and sustaining vulnerabilities (Goldring et al. 2009). In the next section, we briefly highlight the role of administrative discretion as an example of such actions.
Fineman first proposed an alternative approach to human subjectivity by employing the notion of vulnerability as an alternative to legal (and especially formal) models of equality, which continue to dominate contemporary human rights theory (Baumgärtel 2020, p. 14): "I want to claim the term "vulnerable" for its potential in describing a universal, inevitable, enduring aspect of the human condition that must be at the heart of our concept of social and state responsibility" (2008, p. 8). Her conceptualization of vulnerability is founded on the idea that most events that may render us vulnerable are often outside of an individual's control (i.e., injuries, climate disaster). At the individual level, vulnerabilities will range in magnitude because of our positionalities. Nonetheless, vulnerability for Fineman is a shared human condition and not simply an individualized characteristic. A vulnerability analysis, she adds, must focus on both individual and institutional components and overall eschew ideas of a "liberal" (which privileges autonomy and self-sufficiency) for a "vulnerable" subject (p. 11). Fineman concludes that we need to ". . . recogniz[e] that autonomy is not a naturally occurring characteristic of the human condition, but a product of social policy" (p. 23). She calls for society to "mediate, compensate, and lessen our vulnerability through programs, institutions, and structures" (Fineman 2008, p. 10). This call for change, and the move away from the liberal legal subject, is at the heart of Fineman's (2008) writings.
At a broader level, Fineman's (2008) vulnerability theory also calls for a "post-identity paradigm" (p. 17) and a move away from identity-based social movements. Since economic vulnerabilities can be experienced by all people and vulnerability is universal, her model proposes that society move away from organizing around race, gender, or other identities, and instead, organize around broader, shared human vulnerabilities. However, such an approach risks minimizing the crucial efforts and importance of identity-based social movements organizing around racial justice or trans* rights, to briefly name only two concerns. Scholars have rightly criticized her theory for its failure to understand the reality of racial profiling and the role of privilege more generally (Cooper 2015), as well as her reliance on resilience, which risks reinforcing a ". . . neoliberal logic of individualized self-management" (Davis and Aldieri 2021, p. 321).
While there is no question that acts of essentialism and paternalism by decision makers ought to be identified and squashed (Baumgärtel 2020), calls for post-identity analyses of vulnerabilities and/or post-group movements based on these risks are equally fraught with peril. When decision makers are called to ignore group or identity-based vulnerabilities, they may misunderstand evidence that could otherwise lead to a positive decision. For example, refugee adjudicators require informed understanding of the specific issues or vulnerabilities that may impact migrants with diverse sexual orientations and gender identity expressions in their home countries, such as deciding not to report crimes against them to the police or family members. Our position is not that a person's identities create an inherent vulnerability, but rather it is the intersection of specific identities or social conditions with that of social processes, such as regressive state laws or community violence, that creates a specific precariousness or vulnerability. Being informed of these vulnerabilities (e.g., both on an individual level as well as more broadly on a social subgroup level) may allow the decision maker to better understand a person's lived experience, as well as more appropriately analyze credibility and evidence (Cameron 2018). These misunderstandings can be made worse if they have either too little or not enough discretion to gain a better understanding of the identities in a specific context to make fully informed decisions, if their decision making is not transparent enough, or cannot be fully reviewed upon appeal. Being informed by group or identity-based categories does not negate the ability for decision makers to take an individual or an intersectional analysis, as they may also consider how income levels, race, disabilities, or religion may further intersect with sexuality and gender expression (Crenshaw 1991). In sum, being mindful of the construction of vulnerabilities in an immigration law context is not just significant theoretically, but it may shape the training and education of individual decision-makers that can, in turn, impact their subsequent credibility assessments, as we will demonstrate in our selected examples.
Irrespective of these criticisms, scholars continue to take up and refine vulnerability theoretically, including in the migration context, for instance in relation to human trafficking in Brazil (Pérez 2016). As Baumgärtel's (2020) recent analysis of jurisprudence of the European Court of Human Rights (the ECtHR), has shown, the concept has also been at the heart of landmark rulings in the area of migration, albeit with inconsistent legal outcomes. In response, Baumgärtel offers up the concept of "migratory vulnerability," which he defines as ". . . a cluster of objective, socially induced, and temporary characteristics that affect persons to varying extents and in different forms." (p. 13). In other words, from the perspective of migratory vulnerability, vulnerability should be identified only on a case-by-case basis and always in relation to an identifiable social process (Baumgärtel 2020).
Other migration scholars have gravitated towards the concept of "precariousness" instead. The reasons for this move are two-fold. First, the concept places a stronger emphasis on the construction of vulnerabilities by state policies, laws, and social processes (Goldring et al. 2009;Atak et al. 2018). Furthermore, precariousness captures a broad range of irregular and illegalized statuses in law and is often highly gendered and racialized (Goldring et al. 2009). It also results in precarious access to services, for example health care. Secondly, a focus on precariousness moves us away from the negative connotation of vulnerability, which frames individuals as helpless victims or centers individual failure. While we find that the concept of precariousness is better suited theoretically because it more accurately captures the active role of the state, for the purposes of this paper, we use both precariousness as well as vulnerability in our analysis, since the term precariousness is less frequently used in government documents or in case law. We further avoid referring to migrants as inherently vulnerable and instead opt for "migrants made vulnerable" to help signify the external processes at play. As Atak et al. (2018) put it, a framing of precarity acknowledges the external and systemic production of conditions that create vulnerable situations that place individuals at risk. Trends suggest that precarious statuses will likely continue to expand (Goldring et al. 2009), further underlining the constructed nature of the concept and the role that states play in sustaining it. One way in which state laws, policies, and administrative logics may produce and perpetuate vulnerability is through the opaque exercise of administrative discretion or through missing rationales for differential treatment.

Discretion, Transparency, & State Power
Scholarship has frequently analyzed the interwoven relationship between law, discretion, and the production of state sovereignty, particularly in the migration context. Immigration officers (or "street level bureaucrats," (Lipsky 1980) undertake "daily exercises in nation-building as they implement ways to determine national identity by deciding who belongs within or outside of the nation state" (Mountz 2010, p. 56). Borders are reproduced and generated by the social relations that operate within local administrative offices, govern exclusion, discursive control, territorial diffusion, and commodify inclusion (Walia 2021). We agree with Heyman (2009) that discretion is not a ". . . formless domain of uncontrollable action", but rather an ". . . analyzable domain of patterned actions that significantly affects law and administration" (p. 367).
In comparing the scope of discretion available to decision makers in Canada and overseas, we draw on socio-legal framings of discretion (Pratt 2010;Pratt and Sossin 2009) that look to eschew approaches that center a law/discretion binary. Unlike in Dworkin's (1977) famous discretion-doughnut analogy, 3 in practice, law is both embedded and overlaps with legal rules (Pratt and Sossin 2009). Drawing on Foucault, Anna Pratt (2005) submits that discretion is a form of positive social power or a technique of government that works to enable certain forms of government in various settings. Likewise, Chowra Makaremi (2009) notes that discretion is not simply a lack of regulations or a loophole in the law, rather, discretion ". . . is a technique to manage migration flows in real time with the purpose of confining or displacing bodies" (p. 414).
In our analysis, we show that overseas resettlement officers enjoy a much wider and opaquer range of discretion in comparison to their in-Canada counterparts, IRB members. These practices of discretion within resettlement procedures allow the Canadian state they represent to govern and reproduce state sovereignty-or inclusion and exclusion-in ways that are not always clearly justified or open to re-examination.
Before moving on to our close reading of vulnerability, we provide some background regarding the evolution of the two very different "worlds" of Canada's refugee protection system-the inland asylum (refugee) determination process and the overseas resettlement system. Both streams have been subject to significant academic discussion and critique over the decades that we cannot adequately reproduce and summarize here (e.g., Rehaag 2019; Cameron 2018; Labman 2019; Macklin 2005). For our purposes, it is important to note that these two components did not used to be separate streams, but only evolved to be such since the 1980s. The short historical excursion further underlines the power of the state to control the scope and nature of who is worthy of Canada's protection. As we will see later on, the concept of vulnerability has become more closely associated with Canada's commitment to humanitarianism, which is most closely linked to its voluntary resettlement stream compared to the fulfillment of its legal obligation under the 1951 Convention, as now represented by its inland refugee determination system.

Canada's Approach to Protection: The Evolution of Two Different Systems
When considering the two components of Canada's contemporary approach to refugee protection, the inland asylum and the overseas refugee resettlement systems both took shape and operated for years before there was a statutory basis for them, as Canada did not become a party to the 1951 Refugee Convention until 1969 and did not incorporate the Convention definition of a refugee into domestic law until the 1976 Immigration Act. Moreover, Canadian refugee policy has always been influenced by its larger immigration policy (Kelley and Trebilcock 1998) and by its geo-politically privileged position, which made it difficult for refugees to reach Canadian territory. Until the mid 1960s, due to Canada's "white settler" mentality, new immigrants to Canada came almost exclusively from Europe, even those who came because they sought protection from persecution, for example, thousands of Hungarians and Czechs in the 1950s and 60s (Anderson 2013). Further to that, these early refugees were admitted exclusively from overseas, using executive discretion, that is, via ad hoc humanitarian admission schemes, which afforded them only limited rights (Soennecken 2013).
For several decades afterwards, Canadian policymakers expected that most refugees would continue to be admitted to Canada exclusively via overseas resettlement (Labman 2019, pp. 38-40). At the time, government officials used resettlement strategically to respond to specific humanitarian crises of their choosing, for instance by admitting 7500 Ugandan Asians in 1972, the first non-European refugees (Labman 2019, p. 35). Moreover, the government implemented refugee resettlement in such a way that candidates for government-sponsored resettlement were expected to meet labor market criteria similar to other types of immigrants ("ability to successfully establish" criteria) (Casasola 2001).
Canada's private sponsorship of refugees first developed in parallel with government sponsorship during the postwar years. As Cameron (2020) notes, it was religious groups seeking to bring refugees to Canada, who did not meet labor market criteria that initially led to ad hoc arrangements, which allowed them to privately sponsor refugees for resettlement in Canada (33). During the 1950s and 1960s, the Canadian government used private sponsorship to resettle religious minorities and those fleeing political repression from communist countries (Cameron 2020, pp. 40-45). The program then expanded rapidly during the 1970s when 60,000 refugees from Vietnam, Laos, and Cambodia came to Canada (Molloy and Simeon 2016). The contemporary version of Canada's resettlement program has a much more modest intake-government admission targets for all resettlement streams combined in 2019 have been steadily between 25 to 27,000 (Immigration, Refugees and Citizenship Canada 2019).
Canada did not develop an inland asylum system until the 1960s (Anderson 2013(Anderson , 2010. Until this time, inland claims were considered solely on ministerial discretion. The first incarnation of an inland asylum process took root in 1967 under the Immigration Appeal Board, a tribunal tasked with deciding refugee claims. Until 1974, the IAB focused on reviewing paper-based applications at the first instance, scheduling oral hearings only for claims deemed likely to succeed. It is unclear what criteria the IAB used to render decisions for most of this period, as the Refugee Convention was not written into Canadian law until 1973. Failed applicants were able to seek judicial appeal on points of law at the Federal and Supreme Court provided they could obtain leave. From 1974 to 1989, first-instance refugee claims were decided by the Advisory Committee on Applications for Refugee Status (RSAC), which initially continued the approach of its predecessor, with the IAB now providing administrative appeal for failed applications (Anderson 2010, pp. 946-47). After the amended version of the Immigration Act came into effect in 1978, the RSAC began reviewing transcripts of interviews conducted with refugee applicants, while a special review committee would consider failed claims on humanitarian and compassionate grounds at the first instance (in addition to the administrative appeal conducted by the IAB and judicial appeal before the Federal and Supreme Courts).
Despite these developments, until the formation of the Immigration and Refugee Board of Canada (IRB) in 1989, all refugee applications were subject to ministerial approval and any procedural rights granted to applicants were based on a purely discretionary basis. The creation of IRB stemmed, in large part, from Singh et al. v. Canada (Minister of Employment and Immigration), [1985] 1 S.C.R. 177, in which the Court held that refugee claimants physically present in Canada were entitled to an oral hearing. Under the current system, refugee applicants are guaranteed an oral hearing at the first instance where their claim is decided by the Refugee Protection Division (RPD) of the IRB. Failed applicants may appeal decisions made by the RPD to the Refugee Appeal Division (RAD) of the IRB, which was added in 2012 and has discretion to issue a decision with or without a hearing or instruct the RPD to re-assess the claim. Furthermore, as under the previous system, failed applicants may seek judicial review pending court approval. 4 The framework of Canada's asylum system largely remained consistent during the 2000s and 2010s, but a series of successful and failed reforms have effectively reinforced the structure of this process. Chiefly, the Conservative Harper government introduced a series of reforms including (2006 to 2015) new deterrence measures, such as the expanded use of detention, denying applicants access to Canada's public healthcare system, and the creation of a two-tier application process through which applicants on a designated safe third country list would be granted a first-instance hearing, but not the right to appeal negative decisions. Though in effect for several years, some of the key elements of the Harper government's agenda were repealed by the subsequent Liberal government under Trudeau or struck down by the courts. Many continue to remain in place (Anderson and Soennecken 2018).
While civil servants dominate both the overseas refugee resettlement and inland asylum systems, as we will see, resettlement decisions are made in an opaque manner by IRCC immigration officers whose training and routines are largely obscured from the applicants, refugee advocates, and most scholars. 5 By contrast, asylum claims are reviewed by the IRB's RPD members whose training and routines have been studied by several scholars with varying degrees of scrutiny through document analysis, interviews, and participant observation (Cameron 2018;Rehaag 2008;Tomkinson 2018;Rousseau et al. 2002). Furthermore, poor and inconsistent decision making has attracted both academic and media attention, leading to greater public awareness of their activities and to a change in appointment criteria (Rehaag 2019).
Another key difference between the overseas refugee resettlement and inland asylum systems is an asymmetry of procedural rights. Overseas resettlement decisions are conducted by immigration officers whose rationale for accepting or rejecting applications formally follows the "protection" criteria spelt out in the Immigration Act and its regulations, but, in reality, the criteria remain opaque since their interpretation is based on discre-tion. Decisions are also not published and only rarely judicially reviewed (Thériault 2020), which is reinforced by the fact that applicants for refugee resettlement have little access to legal counsel. In other words, while UNHCR and private sponsorship groups may advocate on their behalf they are not guaranteed representation, let alone resettlement, in the overseas decision-making process. By contrast, asylum applicants in Canada have the right to speak directly with RPD members and confront evidence presented to justify rejecting their applications, a crucial aspect of procedural fairness that is guaranteed to all inland asylum applicants at the first instance. Further to that, while legal representation is not provided, at least some legal aid continues to be available and recognition rates before the IRB continue to remain high compared to other advanced industrialized countries. Moreover, rejected applicants are able to have their decisions reviewed by an appeal branch of the IRB, the RAD, and beyond that, by the federal courts, if granted permission. Moreover, although not all of their decisions are published, RPD members are provided with publicly accessible guidelines and must write formal decisions grounded in Canadian law clearly explaining their rationale for accepting or rejecting the claim. Putting aside the conduct of RPD members, asylum applicants have considerably more procedural rights compared to refugees being considered for resettlement overseas.
These two different refugee admission systems and accompanying bureaucratic logics have been in operation since the late 1980s, as mentioned before, and have faced their share of criticism over the years. However, what is interesting to note for our purposes is that only once did a prominent government report, Not Just Numbers (Immigration Legislative Review 1997), suggest reverting back to a single process for deciding both inland asylum as well as overseas refugee resettlement applications using identical assessment criteria. While the report was eventually ignored by policymakers for a number of reasons, it is one of the few that considered processing refugees who had already made it to Canada using the same criteria as those still overseas. 6 If the overall goal for Canada's refugee policy is inclusion, then stronger justifications and rationales for differences between the legal categories and procedures in the in-Canada refugee determination system compared to its overseas resettlement stream, especially with respect to the employment of the concept of vulnerability, should be developed.

Searching for Common Ground: Migrants and Vulnerability in Canada and Overseas
With this section, we begin our close reading of Canada's laws and regulations concerning vulnerability. Fundamentally, looking at the IRPA and its accompanying regulations (IRPR), it is notable that the term "vulnerable" is only referenced in one section of the regulations (IRPR s.138)-and that is with respect to Canada's overseas refugee resettlement program. There is no equivalent in either the legislation or the regulations for in-Canada claims. As we will see, this small imbalance is symbolic of a larger imbalance. Nonetheless, the IRB's Chairperson may issue guidelines to the members of the Board (i.e., its adjudicators) to help them with their duties, as well as identify certain decisions as jurisprudential guides (IRPA s. 159(1)[h]). It is at this level that we find references to vulnerability. 7 However, while members must consider the various guidelines, they ultimately have the discretion to determine if a guideline is applicable or not, which we will address later. As we will discuss after that, that there is no publicly available equivalent to these guidelines in the refugee resettlement program administered by IRCC. While there are procedural manuals and operational guidelines, none contain references to vulnerability.
We have, therefore, opted to discuss the IRCC's urgent and special needs designations, which come the closest to detailing what the label of vulnerability means in the overseas resettlement process.

In-Canada Procedures: The IRB Guideline on Vulnerable Persons and Designated Representatives
In the IRB's (2006) Chairperson Guideline on Vulnerable Persons ("Procedures with Respect to Vulnerable Persons Appearing Before the IRB," Guideline 8), first released in 2006 (Cleveland 2008, p. 119), 'vulnerable persons' are described as: . . . individuals whose ability to present their cases before the IRB is severely impaired. Such persons may include, but would not be limited to, the mentally ill, minors, the elderly, victims of torture, survivors of genocide and crimes against humanity, women who have suffered gender-related persecution, and individuals who have been victims of persecution based on sexual orientation and gender identity. (IRB, Guideline 8, s. 2.1, our italics) An individual may be identified as a vulnerable person by either a Board member or by their counsel at any point in the process. What stands out is that the examples in the list above are a mixture of personal and situational characteristics and that the emphasis is on "severe" impairment. The guideline overall references the term "vulner*" 68 times. Various other terms are used throughout including: 'traumatized', 'survivors', and 'victims.' Nevertheless, while the Guideline acknowledges that all refugee claimants appearing before the Board may have some vulnerabilities, Guideline 8 is explicitly limited in the scope of its applicability: This guideline addresses difficulties that go beyond those that are common to most persons appearing before the IRB. It is intended to apply to individuals who face particular difficulty and who require special consideration in the procedural handling of their cases. It applies to the more severe cases of vulnerability. (IRB, Guideline 8, s. 2.3; our italics) While Guideline 8 is the primary guide used by Board members for accommodating vulnerabilities and, ultimately, aims to ensure procedural fairness, critics have long highlighted that it appears to only be used very infrequently (there is currently no public data on these requests), and that despite the low costs to Canada, there appears to be little willingness to expand the accommodations offered (Cleveland 2008, p. 122), which may include any of the following: a.
allowing the vulnerable person to provide evidence by videoconference or other means; b.
allowing a support person to participate in a hearing; c.
creating a more informal setting for a hearing; d.
varying the order of questioning; e.
excluding non-parties from the hearing room; f.
providing a panel and interpreter of a particular gender; g.
explaining IRB processes to the vulnerable person; and h.
allowing any other procedural accommodations that may be reasonable in the circumstances. (IRB, Chairperson Guideline 8, s. 4.2) The limited use of Guideline 8 has also been confirmed by case law. As Purkey (2022) (this issue) notes, in Hernandez v. Canada, the Federal Court found that ". . . a departure from the Guidelines which does not result in a breach of natural justice, or a breach of fairness would not necessarily give rise to independent grounds for judicial review." Likewise, in Hurtado v Canada, the Federal Court confirmed that vulnerability operates on a spectrum or hierarchy, in line with the philosophy expressed in Guideline 8 (see further case law discussion in Cameron 2018).
Legal precedent further demonstrates that the Board tends to be strict or inconsistent in applying determinations of vulnerability, Purkey adds. For example, the Federal Court ruled in Hilary v. Canada that the Board is not obliged to investigate whether the claimant understands the refugee status determination process even if said person has been diagnosed with mental illness, which further affirms decision-makers' discretion to assess vulnerability. Moreover, even in the presence of seemingly strong proof of psychological conditions and markers of vulnerability, the Board has full discretion to deny such recognition. As an example, Purkey (2022) (this issue) highlights Gardner v. Canada, a case where an applicant before the Board was diagnosed with schizophrenia yet not found to be a vulnerable person. In another case overturned by the RAD, the RPD had applied an overly narrow definition of vulnerability when assessing the application of an unpresented woman claimant with limited education who was a survivor of gender-based violence (RAD No. TB4-04669).
Finally, while Guideline 8's use in identifying and accommodating vulnerabilities seem rather limited, the Guideline does give the Board a mechanism to allow for a 'designated representative' (DR) per s. 167(2) of the IRPA, arguably the most substantial deviation from regular procedures based on an individual being deemed vulnerable. A designated representative is most commonly used in cases of unaccompanied minors or other persons deemed vulnerable who may be unable to "appreciate the nature of the proceedings" (s. 12.1, Guideline on Vulnerable Persons). A DR can be a family member, friend, or an employee at a social service agency that offers DR support. The role of a DR can vary; however, they often support the claimant in instructing counsel, making decisions, or assisting in the decision making regarding the case, assisting in evidence collection, and, ultimately, assisting the claimant understand each procedure before the Board. The DR does not replace counsel, but if the claimant has not secured counsel, the DR has the authority to determine if counsel should be secured. Ultimately the Board members decide whether a DR will be appointed, nonetheless, they must submit a procedural fairness letter to counsel explaining their concerns with the request (if said request for a DR was placed in writing), otherwise they may find their decisions subject to judicial review (s.166 IRPA). While it is currently unknown if securing a DR in a claim before the Board increases the likeliness of a positive decision for vulnerable adults because of the lack of publicly available data, the broader literature on legal representation in refugee determination hearings in Canada would certainly suggest that any legal representation may do so (Tomkinson 2014;e.g., Rehaag 2011).
In this section we discussed the Board's Guideline on Vulnerable Persons, some relevant case law, and the role of the DR. A positive element of Guideline 8 is that it provides members with the ability to accommodate procedural accommodations not listed in the guideline itself, allowing members and counsel to think outside the box in terms of what would be most appropriate for a claimant identified as having vulnerabilities. The Guideline also maintains a broad definition of vulnerability, allowing it to be applied to claimants who may have vulnerabilities not specifically addressed on the Guideline itself.
Yet our brief excursion into the case law highlights the contrast between law in books and law in action: considering the inconsistent nature in which the guideline is applied in practice, with some members applying stricter criteria than others, the potential for discretion to perpetuate or worsen vulnerabilities looms large. Additionally, it appears as if members are not required to explain in their decision how they applied a specific guideline, often indicating simply that the guideline was referenced. We speculate that indicating that a guideline was consulted in proceedings where its application is not entirely evident, may be a tool used by some members to avoid judicial scrutiny.
Returning to our discussion on discretion it is important to reiterate that our analysis of case law is limited to cases that have been appealed, as first instance RPD cases are not publicly accessible. Our focus and concern of discretion, therefore, is not directed to a few individual members' decisions, but ultimately concerned with the specific procedures and mechanisms of the Board that allow for actions and decisions which may worsen precarious situations or, ultimately, lead to the rejection of claimants and potential future citizens. We recommend establishing a process whereby Board members would be invited to detail how guidelines were consulted in detail. This may provide a better level of transparency in the Board's decision making.
Collectively, Guideline 8 and the DR are the most immediately related procedures explicitly designed to address and accommodate identified vulnerabilities for in-Canada claims. We use the language of "identified vulnerabilities" here since requests for procedural accommodations or for a DR by counsel or the claimant themselves are not always granted. In these processes, Board members use individual or context sensitive analyses in their understanding of vulnerability, and temporally speaking, are based on present needs. Fundamentally, this individual level analysis is not about understanding vulnerabilities as individual flaws, but is a means to procedurally support claimants throughout the hearing so that their current situational needs do not interfere with their ability to present as credible. As we will now show in our overseas comparison, vulnerability in resettlement serves a completely different purpose.

Overseas Resettlement-General Principles
When it comes to Canada's resettlement program, the use of identifying vulnerabilities serves a very different purpose compared to the inland process. Whereas vulnerability is treated as an individual and/or situational characteristic by the IRB, which may then lead to procedural accommodations only, 8 the IRCC employs the concept of vulnerability as a tool to help identify and prioritize cases according to its various resettlement categories (UNHCR 2018a).
More specifically, in their documentation, both the IRCC and UNHCR use the language of vulnerability and urgency interchangeably. If a case is identified as urgent and the individual as vulnerable, it will be processed more quickly. For example, someone who is at imminent risk of being kidnapped or executed because of their political views would be considered in urgent need of resettlement because of "immediate threats to life, liberty or physical well-being." Moreover, some assessment criteria may be waived or interpreted more flexibly by the officer, chiefly, the criteria that individuals must demonstrate an ability to be self-sufficient within a few years after resettlement. In other words, refugees who may otherwise not qualify for resettlement to Canada may become eligible if their case is identified as urgent or vulnerable. Other potentially exclusionary criteria, for example, pertaining to high health and medical needs, may also be waived (UNHCR 2018a) for "urgent" or "special needs" cases.
With respect to special needs, IRCC officers can flag certain vulnerabilities to better identify specific resettlement needs (called "special needs" in the IRPR) from the outset and subsequently have these individuals matched with services in Canada after arrival where possible, which is a significant contrast to the in-Canada procedure. 9 Special needs can range from accommodating a large number of family members to trauma, disabilities, and "the effects of systemic discrimination" (157(1) and (2) IRPR).
Interestingly, identifying vulnerabilities does not appear to warrant any formal procedural accommodations for the critical interview with IRCC. Unfortunately, there is no publicly available information to what extent UNHCR field staff employ discretionary measures to accommodate refugees in their screening and assessment prior to any interviews with Canadian officials beyond those developed for individuals in particularly vulnerable 8 For unaccompanied child refugees, the respective guideline urges faster processing. Individuals identified as vulnerable in the detention process are more likely to have their detention conditions reviewed sooner or to be eligible for alternatives to detention, as per the respective guideline. 9 On the government side, identifying such "special needs" then triggers them being slotted into a different resettlement category (Joint Assistance Sponsorships or JAS). Individuals in the JAS program are explicitly labelled as "in greater," "special" or "exceptional" need in government documentation. They are brought to Canada in collaboration with a private sponsorship group. There are no publicly available statistics for this category. A 2003 estimate concluded that up to 10 percent of government sponsored refugees are JAS cases. (Treviranus and Casasola 2003, p. 191; Government of Canada 2018).
circumstances (e.g., encampment or detention), underlining the potential for discretionary to become murky and problematic (see also Flegar 2018, p. 377). 10 Overall, while there are specific vulnerabilities discussed within the UNHCR Canada Resettlement Handbook, there are not always equivalent categories for these same refugee claimants in the Canada Resettlement Handbook (i.e., older refugees). These broad differences in categorization and varying treatment of persons with 'vulnerabilities' by the IRB for in-Canada claimants compared to Canada's overseas resettlement program raises concern on ethical and normative grounds. 11

Approaches to Identifying and Accommodating Vulnerability Overseas
Compared to the in-Canada procedure, in the Canadian overseas resettlement process, vulnerability is considered at two different stages: first, by UNHCR and then, if they choose to recommend the individual for resettlement, by a Canadian (IRCC) visa officer. Note that there is no counterpart to a "designated representative" in the overseas process. As explained earlier, the Canadian visa officer reviews the UNHCR dossier and may choose to conduct an additional interview with the applicant. If they find that the individual is "vulnerable" or in "urgent need of protection," the case will receive priority processing (on labeling, see generally Zetter 1991). The IRCC's OP1 further states that "refugee applications classified as urgent or vulnerable should [. . . ] receive priority processing" (Government of Canada 2017, p. 17). As mentioned before, it is important to note from the outset that vulnerability is equated with urgency in the Canadian overseas process.
The 'urgent' classification is one of three priority levels used by UNHCR (i.e., emergency, urgent, and normal) in addition to being a frequent adjective and term used to highlight certain vulnerabilities in UNHCR literature. According to the UNHCR handbook, the majority of UNHCR cases recommended for resettlement fall within the "normal" priority level. In contrast, urgent cases will be processed within one to four months, compared to those deemed an "emergency," which will be handled within a seven-day maximum (i.e., persons facing imminent threats to their lives): Urgent need of protection means, in respect of a member of the Convention refugee abroad or the country of asylum class, that their life, liberty, or physical safety is under immediate threat and, if not protected, the person is likely to be (a) killed; (b) subjected to violence, torture, sexual assault, or arbitrary imprisonment; or (c) returned to their country of nationality or of their former habitual residence. (besoin urgent de protection). (s. 138) and s. 139(2) IRPR).
The Handbook's Canada chapter similarly explains that when a case is deemed vulnerable by the IRCC, it is processed and prioritized before regular resettlement cases and within a faster time frame of one to four months (UNHCR 2018a). 12 As already noted, there are no publicly available guidelines for any accommodations offered to such individuals in the IRCC interview similar to the IRB's Guideline 8. However, one key substantive contrast with the IRB Guideline is the fact that IRCC visa officers are required by law to limit all of their vulnerability screening to individuals who are at a severe risk to their physical safety, which restricts the scope of their discretion and, as a result, may leave some migrants vulnerable and in precarious situations: Vulnerable means, in respect of a Convention refugee or a person in similar circumstances, that the person has a greater need of protection than other applicants for protection abroad because of the person's particular circumstances that give rise to a heightened risk to their physical safety. (vulnérable). (s. 138 and s. 139(2) IRPR) 13 In the next section, we briefly review the IRB's Gender Guideline and IRCC's "Women at Risk" program, as well as relevant case law. We are focusing on gender as our next example here because, as Flegar and Ledema caution (2019), though gender, as an example of a visible and embodied group-based label, is frequently included in bureaucratic vulnerability screenings, all too often, it is reduced to "women" who are then all universally labeled vulnerable, which may become stigmatizing. 14 In our discussion, we are, therefore, particularly interested in how vulnerability is understood for this group. As before, we also seek to uncover any differences between the overseas and the inland processes and with what justification.

Gender & Vulnerability in Canada The IRB's Gender Guideline
The IRB's Guideline 4, entitled 'Women Refugee Claimants Fearing Gender-Related Persecution' (the Gender Guideline) was first released in March 1993 and updated in 1996. As Audrey Macklin (1998) notes, it followed a release of guidelines by UNHCR in 1991. Shortly after its introduction in Canada, similar guidelines were introduced in the United States (in 1995) and in Australia (in 1996). 15 It provides the Board with focused considerations for substantive and procedural (evidentiary) matters, including "special problems women refugee claimants experience in demonstrating their claims." The Gender Guideline acknowledges gender-based persecution as a legitimate form of persecution. Despite this, members are still required to identify a connection between the claimant's gender, her feared persecution, and one or more official (or "independent enumerated") grounds of persecution that is recognized under the Convention, since Canada still does not formally recognize gender as a separate ground of persecution for persons seeking status as a protected person in Canada (i.e., in the IRPA), thereby continuing to mirror the 1951 Geneva Convention's definition a refugee-even though the Canadian definition of "protected persons" already goes beyond the Convention. 16 Although the word is not used, it could further be argued that the Guideline also requires that members take an 'intersectional' approach, because they are explicitly asked to consider a claimant's identified gender-related persecution in addition to one or more enumerated grounds listed within the IRPA (similarly, e.g., Aberman 2014, p. 58;LaViolette 2007, p. 187).
Further to that, although the Gender Guideline recognizes that both men and women may face similar types of persecution, they may, nonetheless, have differing degrees of vulnerability. The framework section of the guideline explicitly highlights social 'subgroups' of women, such as their age, race, economic status, or marital status. The Gender Guideline requires members to analyze the cultural and social context in which a woman resides to consider if her 'subgroup' is indeed unchangeable or not. If a member does find that she is a 13 Our italics. 14 Specifically, Flegar and Ledema (2019) warn of problems associated with labelling some migrants inherently as vulnerable. This universal category of victimhood impacts their agency, resilience, and autonomy. It also negates structural factors. 15 Macklin points to the fact that the UK did not see it necessary to introduce similar guidelines, although NGOs had lobbied for them and even introduced their own guidelines (Macklin 1998, pp. 25-26). It took until 2000 for such guidelines to be introduced there (Berkowitz 2000). 16 Article 1A(2) of the 1951 Convention only refers to "membership in a particular social group" as possibly triggering persecution. It is not defined any further in the Convention itself.
member of an identified gender-based social subgroup, this in itself is not enough to satisfy refugee eligibility criteria-she must also demonstrate a genuine fear of persecution. 17 There are four primary concerns flagged as evidentiary matters in the Gender Guideline that members are instructed to review. First, a member cannot dismiss a woman's claim for protection because other women in her country-of-origin face similar generalized violence and oppressions. Second, members are required to consider evidence demonstrating if state protection was available or adequate to protect a claimant from gender-based persecution. In making this assessment, the Guideline indicates that "clear and convincing proof" may not necessarily be available in determining the quality of state protection, as more standard forms of evidence may be unavailable (i.e., statistical data in her country of origin, see IRB 1996). Nonetheless, members are required to consider other forms of evidence, such as a woman's personal testimony, as well as the testimonies of other women in similar positions where state protection from gender-based persecution was unavailable. Third, members are cautioned against equating positive developments in a woman's country of origin with the possibility of a safe return. Members are required to consider that a claimant's fear of persecution may not change in parallel with a country's recent change in circumstances. Fourth, in considering 'internal flight alternatives' (IFA), members are required to assess religious, economic, and cultural factors when considering a claimant's potential move to a new area within her country of origin.
While this is just a basic overview over the IRB process with respect to gender (the literature is vast, e.g., Arbel et al. 2014), for our purposes, it is important to note that the Gender Guideline does not categorize all women as vulnerable from the outset. Only two situations are explicitly labeled with the word "vulnerable" in the Guideline-women who are victims of sexual violence and those who fail to conform to certain discriminatory religious or cultural norms.
Another way to link a woman's fear of persecution to her vulnerability is for Board members to use an intersectional lens or engage multiple guidelines in their decision making at once. For instance, the Board may engage the Guideline for Vulnerable Persons to offer a woman refugee claimant experiencing PTSD procedural accommodations, while also engaging the Gender Guideline to assist in their assessment of her overall refugee claim (Sadrehashemi 2011, p. 8).
However, it is not clear empirically thus far how frequently members do so and whether such a cumulative reading of the Guideline makes protection easier or harder to obtain on average. More research is needed. However, looking to some examples from a recent review of relevant case law (Purkey 2022) provides at least some understanding as to how the Gender Guideline is applied and linked to vulnerability in practice by Board members. While the analysis makes it clear that Board members must take an intersectional approach in their decision making when hearing claims from women refugee claimants identified as vulnerable, claims for protection by women solely based on a generalized risk to gender-based violence in their country of origin were more likely to result in a negative decision.
For instance, several IRB cases involved Haitian women as a social sub-group. Because women are at risk of gender-based violence across Haiti, additional factors also had be present for them to be granted protection, such as being from a 'shantytown' or living with a disability, whereas Haitian women who were not low-income or who were educated or married, did not meet this vulnerable profile and would appear to not qualify for protection (see RAD File No. MB5-01988) (Purkey 2022). Similarly, in cases involving Roma women, they also needed to demonstrate vulnerability to sexual exploitation or human trafficking (e.g., Djubok v Canada) before being granted protection. Purkey (2022) also identified an odd pattern in judicial reviews (i.e., IRB cases which have been appealed in federal court): those involving a failure to apply the Gender Guide-line resulted in more positive decisions compared to those based on failure to apply the Guideline for Vulnerable Persons (see: Ndjizera v Canada, Abbasova v Canada, Bibby-Jacobs v Canada). Purkey (2022) speculates that the reasons for this discrepancy may be that the Guideline for Vulnerable Persons is largely focused on assuring procedural accommodations, whereas the Gender Guideline deals with substantive issues that may have implications for credibility assessment. 18 This finding further supports our suggestion that Board members ought to be required to outline in their judgment how they implemented any relevant guidelines in their decision making, and not simply list that the guideline was referenced. Such broad discretion and failure to properly interpret and apply the Gender Guideline may result in persons fleeing gender-based persecution being denied protection.
The original Gender Guideline is now largely considered outdated in many ways (e.g., it used antiquated language, such as "Battered Women Syndrome"). Arguably, the Gender Guideline also requires more specific examples and situations to help members draw connections between the written guidelines ("the law on books") and its application ("the law in action") that can then be used for training purposes (for a critique of IRB training, read Cameron 2018, p. 34). At the time of writing, the Board has established a new 'Gender Related Task Force' (GRFT), which is composed of 24 Board members who have received specialized training to assist in deciding on gender-related claims (IRB 2020a). According to the Board's 2020-21 Departmental Plan, the GRFT will also establish best practices which will later be shared with the broader RPD (IRB 2020b). Notably, training for new GRFT members lasts four weeks and includes topics such as "trauma informed decision making", "assessing credibility", "cultural humility", and "questioning from a cross-cultural and trauma informed perspective" to name a few (IRB 2020a). In addition to the GRFT, the Board is actively consulting with stakeholders and subject matter experts on the Gender Guideline. The GRFT indeed looks like a step in the right direction for addressing identified concerns with the way in which the Gender Guideline has been applied. Once fully operational, further research into the Board's creation of the GRFT will be imperative to better assess how claimants fearing gender-related persecution are having their claims assessed.
Our analysis of the Gender Guideline, relevant case law, and the creation of the new gender task force helps demonstrate, in part, our earlier claim that first, analyses of vulnerabilities by decision makers does not need to completely eschew group-based vulnerabilities, and second, that intersectional and individual-level analyses can happen in confluence. When hearing claims from persons fearing gender-related persecution, the guidelines lead the Board to apply an intersectional analysis in their decision making by identifying women 'sub-groups'. In addition to an intersectional analysis, the Guideline also serves as an ongoing training tool to remind members of special circumstances faced by women refugee claimants that could impact their credibility assessments. Returning to our discussion of discretion, it is clear from the number of positive decisions resulting from judicial review in cases where the Gender Guideline was applicable, that too much discretion may again work to exclude groups of people who are otherwise legally entitled to protection. This is particularly important when the Gender Guideline engages with special evidentiary concerns, which is highly relevant to credibility assessments, unlike the Guideline for Vulnerable Persons, which is largely procedural.
The example of the Gender Guideline further helps illuminate our position on vulnerability theory and rejection of a "post-identity" or social group analysis. In the context of the IRB, Board members are clearly instructed to undertake a case-by-case and individual assessment of each claimant before them. We have documented that this analysis is also necessarily intersectional. At the same time, members are directed to consult relevant guidelines (e.g., linked to specific identities or situations, such as persons fearing gender persecution or persons with identified vulnerabilities) and failure to do so may result in inappropriate decisions and potentially result in judicial review. The Gender Guideline provides training to members on specific issues (e.g., crucial evidentiary considerations), and works to challenge biases or stereotypes they may have, particularly about migrant women (who are mostly all racialized) fleeing gender-based persecution. Having Board members trained on the potential experiences of women fleeing gender-based persecution in different regional contexts may ultimately assist in their decision making, while being mindful of cultural relativism. It is important to reiterate that the current Gender Guideline is considered outdated and far from perfect; nonetheless, both in theory and in practice, a regularly updated Gender Guideline with broad community stakeholder consultations on its content would be helpful. Notably, the Gender Guideline does not discuss unique issues faced by young unaccompanied men for example, or by trans* refugee claimants. Nonetheless, the newer SOGIE Guideline discussed below expands on the Gender Guideline to include claimants of various sexual orientations and gender identities. As with our first comparison of vulnerability, Canada's overseas special category for gender and Women at Risk serves a different purpose than that of the IRB.

Gender, Resettlement, and Vulnerability Overseas Canada's "Women at Risk" Resettlement Category
Canada first launched the Women at Risk (AWR) resettlement program as a pilot project in 1987 (Madokoro 2018, p. 345). Some government officials involved in the creation of the program had personally witnessed women refugees in encampment situations who had been sexually assaulted or who had been left behind while the men were resettled, often with young children. There was also a sense of desperation, Madokoro adds. Other officials had similarly observed that women without a male head of household were frequently excluded from Canada's resettlement program because they were less able to meet the criteria for successful resettlement in Canada because of traditional gender roles in their country of origin, which, today, is assessed on a sliding scale if these women are not considered vulnerable for additional reasons (UNHCR 2018a).
The contemporary Canadian chapter of the Resettlement Handbook continues to use the absence of a "family" as a key qualification criterium (UNHCR 2018a). More precisely, the Canada chapter now defines 'women at risk' as ". . . women without the normal protection of a family who find themselves in precarious situations and who are in a place where local authorities cannot ensure their safety" (UNHCR 2018a, p. 10). Women refugees who fall within this category could also be identified by Canada as needing additional support with their resettlement after arrival, for example if they are survivors of violence or torture. 19 UNHCR screening, which takes place prior to that from IRCC, relies on two tools to determine eligibility for resettlement, including the Heightened Risk Identification Tool (mentioned briefly previously) and an assessment of resettlement needs based on the 'intensity' of four identified factors (UNHCR 2010). The Heightened Risk Identification Tool is used to ask a series of questions and indicators ranging from "Woman" (including widow, single mother, abandoned older woman, etc.) or "girl without family protection/support" to "Engaging in Survival Sex" (UNHCR 2010, p. 9). Fourteen indicators are checked regarding the claimant's past and present experiences, as well as their family members', which are ultimately scored as part of a final risk assessment, ranging from low to high.
Some women at risk cases may also fall under 'urgent' need. For instance, IRCC's 2018 Annual Report (Immigration, Refugees and Citizenship Canada 2019) highlights Canada welcoming "1400 survivors of Daesh, including vulnerable Yazidi women and children and their families (pp. 9, 25)." All cases identified as eligible for the Women and Girls at Risk program are sent to the Resettlement Operations Division at the IRCC office in Ottawa, where the rationale provided by the local IRCC immigration officer outlining the need for urgency is reviewed.
While Canada's public commitment to prioritizing women refugees remains laudable, this brief look at the very different ways in which vulnerability linked to gender is "operationalized" for in-Canada claims compared to overseas applications underlines the power of the Canadian state to construct very different procedures and legal categories for women refugees who are deemed vulnerable without a clear explanation as to why. In both the in-Canada and overseas stream, immigration officers retain significant discretion to include or exclude women. However, the actions of overseas officers are much more opaque and also much less subject to judicial scrutiny.
In the next section we introduce our final example, the SOGIE Guideline, to expand on the concept of vulnerability and how it is currently operationalized for in-Canada refugee claimants compared to refugees abroad.

Sexual Orientation & Vulnerability in Canada
Sexual Orientation, Gender Identity and Expression, and Sex Characteristics: The IRB's SOGIE Guideline The SOGIE Guideline was first introduced in 2017 to assist IRB members in Canada in better understanding the unique issues and harm individuals may experience when their sexual orientation and/or gender identity does not conform to the cultural norms in their country of origin. The guideline declares that LGBTQI+ refugees (and internally displaced persons) are made vulnerable in countries where people of certain gender identities or sexual orientations are criminalized and/or at greater risk of community violence.
The SOGIE Guideline covers a wide range of topics from appropriate terminology and particular harms individuals with "diverse SOGIE" may experience, to the need to avoid stereotypes and the flagging of important evidentiary concerns. With respect to terminology, the SOGIE Guideline differentiates between gender expression and gender identity, informing members that a person's gender "may be anywhere along the gender spectrum" (s. 2.4, IRB 2021). The guideline goes on to serve educational purposes for Board members, defining terms such as pansexual, bisexual, trans, intersex, cisgender, and queer (s. 2.8-2.9, IRB 2021). It is worth highlighting that where the Gender Guideline does not address trans* persons or other persons fearing gender-related persecution, the SOGIE Guideline has a section (s.8.5.4) dedicated to particular vulnerabilities faced by trans and intersex persons.
The Board is also advised on evidentiary concerns when hearing cases from individuals with "diverse SOGIE". As with cases of women who have experienced gender-based persecution, members are reminded that a claimant's testimony alone may be the only evidence available. When considering country condition reports, the Board is also told to be aware that SOGIE-based violence and related forms of marginalization may be underreported. As with the Gender Guideline, the SOGIE Guideline more explicitly identifies the importance of an intersectional approach: The intersection of SOGIE with additional marginalization factors, such as race, ethnicity, religion, faith or belief system, age, disability, health status, social class, and education, may create both an increased risk of harm as well as distinct and specific risks of harm. The intersection of these factors, which are nonexhaustive, may impact an individual's access to state protection or an internal flight alternative (IFA).  Due to the more recent implementation of the SOGIE Guideline in 2017, Purkey's (2022) analysis of case law did not produce many cases (n = 19) with substantive analysis of vulnerability in association with SOGIE. The majority of cases reviewed were decided against the claimant, though Purkey's (2022) highlights this may be the result of sampling as her analysis of case law is limited to appeals at the RAD, the RPD, and judicial review only. For example, in Gorria v. Canada (2004), prior to the implementation of the SOGIE Guideline, the Board found that "working-class homosexuals are more vulnerable to harm." Similar to the case of the Haitian women, who only qualified for protection because they were also low income, this claimant was deemed to not fall within this economic category and was not offered protection.
A recent (2021) decision at the RAD, which found in favor of the claimant, highlights the risk of discrepancies between the SOGIE Guideline in its written form and its application in practice by individual Board members (RAD file No. MB9-29401). In their review of the RPD's negative decision of a young, gay male from Algeria (i.e., a country where same-sex acts are criminalized), the RAD found, ". . . the RPD incorrectly drew a negative inference about the appellant's credibility on the basis that he did not seek asylum earlier," (translated from French) and added: Not seeking asylum at the earliest opportunity may appear incompatible with a well-founded fear of persecution for a non-vulnerable person who is familiar with the system of protection that Canada offers to refugees, but it is not the perspective of a newly arrived survivor of sexual violence in Canada. . . I am of the view that the RPD has shown a lack of compassion towards him. . . The RPD did not consider, in analyzing the appellant's testimony, the impact that this trauma might have on his memory or on his ability to testify to the elements at the source of this trauma. (RAD file No. MB9-29401, translated from French) The RAD drew a positive inference between the claimant being a young, gay male, who was also a survivor of sexual violence and newly arrived migrant to Canada with being a vulnerable person and in a precarious situation due to external factors. The RAD goes as far to say that the RPD member lacked compassion and that ". . . some of the RPD's questions are inappropriate or even based on stereotypes. For example, on several occasions the RPD asks his questions as if the relationship between the appellant and his attacker were consensual" (RAD file No. MB9-29401, translated from French). While deeply troubling, RAD file No. MB9-29401 highlights the dire need for ongoing training of the SOGIE Guideline on a regular basis to ensure members are consistent in their application. Returning to our discussion on vulnerability theory, this example further emphasizes the need for decision-makers to be trained specifically on group-based identities (e.g., realities for LGBTQI+ migrants). Yet increased training and knowledge does not take away from the need for an individual-based or an intersectional analysis in decision making. Moreover, an intersectional approach should not be used as a tool against the claimant, i.e., parts of someone's intersectional identity should not be used to make it harder for them to obtain protection. The ultimate question is not simply how well any SOGIE Guideline is written (i.e., how good the "law on books" is constructed), but how well members understand and apply it in their decision-making (i.e., the law in practice). Purkey (2022) further notes that the case law also referenced identifications of vulnerability by international sources. For example, in Melchor v. Canada, the Mexican government's statement that gay people are a vulnerable group is referenced. Likewise, in Montero v. Canada, an Ombudsman in Costa Rica is referenced stating that people with "distinct sexual preferences" are a vulnerable group. However, the impact of these assessments on an individual's chances to obtain protection is empirically unclear.
In their time spent volunteering, 18 months with two LGBT refugee support groups, David Murray (2020) found that: The applicants were reminded repeatedly by their lawyers, peer support group leaders and each other that there were a number of components, characteristics and assumptions utilized by IRB Members to determine the credibility of a SOGIE refugee claim, and that if they learned and understood these assumptions and characteristics associated with "LGBT" identities, and integrated them into an appropriate narrative of identity formation and persecution based on that identity, then they stood a better chance of a successful hearing. (p. 72) Murray (2020) posits that when hearing claims from SOGIE refugee claimants, that the Board has ". . . assumed understandings of 'real' or authentic gay, lesbian, bisexual or transgender identities that reflect primarily white, middle-class experiences and beliefs about those sexual and gendered identities. . . " (p. 74). In concluding, Murray (2020) argues that these processes (IRB hearings and SOGIE Guideline application) produce a form of homonationalism, which works to decide which migrants belong and which will be a excluded for not fitting the ". . . narrative's acceptable performances, characteristics, and/or aesthetics" (p. 74). These findings may support our first suggestion that like the Gender Guideline, the SOGIE Guideline should undergo regular review with wide input from community stakeholders. It is particularly important for these guidelines to include many detailed examples, as they also serve as an educational and training tool for members. Second, when Board members do not have to record how they implemented specific guidelines in their decision making, and simply state that they referred to a specific guideline, this type of discretion may operate as a technique of government and allow for instances of national gatekeeping. We hope that our interviews in phase two of the VULNER project will shine more light on this issue.
The SOGIE Guidelines were recently updated following 11 recommendations that stem from the IRB's review of the SOGIE Guideline (IRB 2021). 20 The accepted recommendations include voluntary paid SOGIE training for interpreters, adding a SOGIE component to newly hired personnel, and updating training for designated representatives to include reference to the SOGIE Guideline. Another recommendation would require members to undergo training every time the SOGIE Guideline is updated. A final recommendation calls for a fulsome review every 3 years. There are, however, no specifics regarding the length of training or the rate at which training will be repeated for members beyond the times at which the guideline is updated. Given that the SOGIE Guideline, as with the Gender Guideline, deals with special issues and evidentiary concerns and goes beyond procedural accommodations, it is surprising that there is no annual training on these guidelines given their importance and relevance to credibility assessments. Nonetheless, the SOGIE Guideline update promises to ". . . provide additional examples of approaches to assist members in identifying their own biases and assumptions, as well as appropriate ways to assess credibility" (IRB 2021). In our next and final resettlement example, we explore how persons identifying as LGBTQI+ overseas have their applications processed and note the mass differences between the in-Canada and overseas procedures.

Searching for LGBTQI+ Individuals in the Resettlement Context
At the time of writing there is no explicit mention of sexual orientation or LGBTQI+ identification as a ground of protection or as a particular vulnerability in the Canadian chapter of the UNHCR resettlement handbook, perhaps because it was last updated in 2018. There are also no publicly available guidelines that Canadian visa officers may use to assess and prioritize any LGBTQI+ persons referred to them by UNHCR for resettlement to Canada. This is surprising, given that the larger UNHCR handbook dedicates an entire section to their protection needs (UNHCR 2018b, p. 199) and that Canada is home to several NGOs dedicated to privately sponsoring and resettling LGBTQI+ individuals. 21 What is more, LGBTQI+ advocates have highlighted that training for UNHCR staff in the field-who are responsible for screening, interviewing, and, ultimately, recommending 20 The revisions were released on 17 December 2021. 21 We are aware of Canadian (and American) private sponsorship initiatives specifically aimed at resettling LGBTQI+ individuals, for example Rainbow Railroad. Rainbow Railroad is advocating for a dedicated resettlement stream for LGBTQI individuals and/or them to be eligible for the UPP program, which is currently only open to individuals who have crossed a national border. For more, read https://www.rainbowrailroad.org/the-latest/three-steps-the-next-canadian-government-can-take-toprovide-more-pathways-to-safety-for-lgbtqi-refugees ( LGBTQI+ individuals for resettlement-varies greatly by country, despite recent funding for additional training received, for instance, from the United States government. More concerning are reports that local staff working for UNHCR and other aid agencies in some field offices continue to discriminate against LGBTQI+ refugees (Fisher 2019, p. 289) and that LGBTQI+ individuals are often concerned about privacy and data protection, which leads to delays in their registration with UNHCR, in particular in urban areas (Heartland Alliance 2014; cited in Fisher 2019). As Fisher notes, UNHCR also does not currently track LGBTQI+ individuals separately and resettlement countries do mostly not offer consistent procedural accommodations, such as access to an interviewer of a preferred gender or access to counsel. Clearly, more research is needed in this area. As far as we can tell, LGBTQI+ individuals who want to be resettled to Canada, have to meet one of the Canadian government's resettlement categories-unless they are privately sponsored. 22 The most likely candidate is UNHCR's legal and/or physical protection needs category, which we now briefly review before turning to our conclusion.

Vulnerability and the Legal and or Physical Protection Needs (UNHCR Category)
Persons considered for resettlement under UNHCR's Legal and/or Physical Protection Needs category must meet one of three criteria, including: threat of refoulement to their country of origin; threat of arbitrary arrest, detention, or deportation; and/or threat to physical safety or fundamental human rights in their present country of refuge. Common vulnerabilities listed in this category include persons who are: 'LGBTI', persons experiencing gender-based violence (i.e., trafficking for sexual slavery, domestic violence, or forced marriage), or persons caught in violent family feuds in the current country of refuge (UNHCR 2018b, p. 198).
For persons seeking eligibility under this category, UNHCR states that a history of harassment is not sufficient to qualify for this category; an eligible refugee for this category must have a present and imminent threat in their country of refuge. It is interesting to note that under this category, UNHCR field staff are required to demonstrate that efforts have been made to resettle refugees internally within their current country of refuge. Only when internal resettlement has been considered or attempted and it is not deemed possible can a person then be considered for this UNHCR category.
While this category offers governments the option to extend protection to vulnerable LGBTQI+ individuals, the "imminent threat" requirement, together with need to demonstrate attempts at internal resettlement, limits protection compared to applicants who have already made it to Canada, as is evident by the comparatively more expansive in-country guidelines.

Conclusions
This article analyzed how Canadian laws and regulations have constructed two fundamentally different worlds of protection for migrants deemed vulnerable-one open only to those who have managed to make it to Canada and who are seeking protection there, the other open only to individuals who are being considered for resettlement to Canada while still overseas. While the development of these two worlds can certainly be linked to the historic evolution of Canada's two different refugee streams and, perhaps more so, to the transformation of the resettlement stream into a strategic, humanitarian tool by UNHCR from the early 1990s on, it remains unclear why there are hardly any similarities between the two streams when it comes to how inclusively and transparently the concept of vulnerability is currently defined in law and in accompanying regulations and how immigration decision makers operationalize it in practice in both worlds when they interact with migrants seeking protection.
From a theoretical perspective, while the Canadian example shows that decision makers try hard to look beyond someone's identity-linked vulnerabilities in their decision making, our brief excursion into the case law in the case of the gender guidelines illustrates that there is a danger that situational vulnerability characteristics are required before protection is granted because someone's identity-linked vulnerability is taken for granted. Clearly, more discussion on how to operationalize intersectionality and how to construct fair and transparent categories is needed.
As we discussed, it is certainly laudable that the IRB has developed a guideline to help its adjudicators to identify vulnerabilities in the context of in-Canada refugee claims as a tool for offering assistance to IRB members in their decision making (chairperson guidelines). This identification serves three purposes. First, most notable, the IRB's chairperson guidelines offer procedural accommodations to claimants who may otherwise face enhanced difficulties in presenting their case. In contrast, there are no comparable guidelines to assist IRCC visa officers in their decision-making process, including with their interviews, at least that are publicly available. The same applies to UNHCR processes, although there are guides for particular circumstances. As Cleveland notes, considering that none of the accommodations explicitly mentioned in Guideline 8 are particularly taxing for the system, at a minimum they should also be offered to overseas applicants in IRCC interview (Cleveland 2008, p. 120). Considering that IRCC and UNHCR interviews serve essentially the same purpose as the IRB hearing in Canada, this would ensure a consistent standard for all applicants seeking protection. Considering the "grave," "significant," "dire", and "potentially even fatal" consequences of decisions being made at both the IRB and IRCC (Cameron 2018, p. 43), it is worth contemplating why some of the procedural accommodations for vulnerable individuals are not already part of the standard process for all applicants before the IRB.
Second, the chairperson guidelines also provide relatively detailed educational information on a series of specific recognized vulnerabilities (i.e., gender-based violence, persons with diverse sexual and gender orientations) that could be the subject of joint training for both IRB employees and IRCC officers. These guidelines appear to serve as a regular educational and training tool for IRB members that include issues or situations that may also be applicable in the overseas context. Finally, some of the chairperson guidelines address specific evidentiary concerns, which have significant implications for credibility assessments that may not be applicable overseas. At the same time, IRCC processes could benefit from prioritizing concerns of physical safety to include those involving mental health. Research with LGBTQI+ individuals suggests that the two aspects are frequently linked.
Finally, it is important to note that while case law has been illuminating in helping to better understand the application of vulnerability and the IRB guidelines, throughout, there are limitations that stem from only being able to review negative IRB decisions that have been appealed in Federal Court, as cases within the IRB are confidential and not readily available for analysis. Moreover, courts are rarely ever called upon to review negative resettlement decisions and if they are, it is mainly due to the Canadian sponsor (Thériault 2020, p. 230). While this creates inequities between government-and privately sponsored refugees, it also illustrates how easily bureaucratic procedures and logics can render individuals seeking our protection vulnerable, leaving them in precarious situations for far too long.
Finally, if Canada wants to retain its leadership role in the global refugee regime and continue to promote its inclusive policies internationally, then stronger justifications and rationales for differences between the legal categories ("law in books") and procedures ("law in action") in the in-Canada refugee determination system compared to its overseas resettlement stream, especially with respect to the employment of the concept of vulnerability, are needed.  (FQRSC, agreement no. 294442). This study is part of the VULNER project, an international research initiative which has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement no. 870845.

Institutional Review Board Statement:
The ethical aspects of the Canadian field research components of the VULNER project were initially reviewed and approved by the University of Ottawa Research Ethics Board (REB file number S-02-20-5463). This project's field research components have also been reviewed by ethics boards from McGill University (REB file number 20-05-020), York University (REB file number 2020-105) and the University of Waterloo (REB file number 43248).

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.