Authored by Yoann Axel Emian, Canadian Immigration Lawyer
I. Introduction
A woman applies for permanent residence on humanitarian and compassionate grounds. She was subjected to sustained violence by her husband in her country of origin. Her family will not receive her back. She has no independent means. He located her twice after she moved; the third time, she left the country.
Those facts constitute a well-founded fear of persecution by reason of membership in a particular social group. They also constitute hardship on return. Not related hardship. The same hardship, described in a different register.
Subsection 25(1.3) of the Immigration and Refugee Protection Act instructs the officer deciding her application not to consider the first and to consider the second. This article argues that the instruction cannot be followed in this class of case, that the class is much wider than intimate partner violence, and that the cost falls on applicants least able to absorb an unreasonable decision.
The point is not that certain crimes are grave enough to earn Canadian protection. Canadian law does not work that way and should not. Gravity is not the test. The test is an individualised assessment of future risk, state protection, the availability of internal relocation and, in the protection stream, connection to a Convention ground. The argument is narrower and more technical. In this category subsection 25(1.3) prevents that individualised assessment from being performed honestly.
II. The rule that produces the problem
Subsection 25(1.3) provides that in examining an in-Canada request, the Minister may not consider the factors taken into account in determining whether a person is a Convention refugee under section 96 or a person in need of protection under subsection 97(1), but must consider elements related to the hardships affecting the foreign national. It was added by the Balanced Refugee Reform Act to stop humanitarian applications from operating as second refugee determinations, which is a legitimate purpose and which the provision achieves across most of the caseload.
The Federal Court of Appeal made it workable in Kanthasamy v. Canada (Citizenship and Immigration), 2014 FCA 113, by distinguishing factors from facts. Matters such as a well-founded fear of persecution and risk to life, being factors under sections 96 and 97, may not be considered. But the facts underlying those factors may be relevant insofar as they relate to whether the applicant is directly and personally experiencing hardship, including adverse country conditions with a direct negative impact on her.
The Supreme Court in Kanthasamy v. Canada (Citizenship and Immigration), 2015 SCC 61, left the distinction intact and opened the surrounding inquiry. There is no hardship test, a point the department accepted in its program delivery updates of 2016 and 2017. The assessment is cumulative, so factors individually insufficient may together warrant relief. Hardship is not exclusively prospective. And the guidance directs decision makers to apply Gender-Based Analysis Plus principles to applications from potentially vulnerable populations, and to consider intersectional identity factors. All of that assumes the factors and the facts can be separated. The rest of this article is about a category in which they cannot.
III. The collapse recurs across every profile
Intimate partner violence, the example in the introduction, is the clearest illustration but not the widest. The groups discussed below are drawn from the Immigration and Refugee Board’s legal services compilation, La jurisprudence sur la définition de réfugié au sens de la Convention et de personne à protéger, current to 31 December 2020, which collects the decisions recognising them.
Female genital mutilation
The compilation records that women subjected to excision have been recognised as a particular social group. UNICEF reported in March 2024 that over 230 million girls and women alive today have undergone the practice, an increase of 30 million or 15 percent since 2016, with 144 million in African countries, 80 million in Asia and 6 million in the Middle East, and a growing tendency to cut girls before their fifth birthday.
In a humanitarian application the analysis fractures immediately. That a woman was cut is completed harm bearing on present physical and psychological condition, which is available. That her daughter has not been cut and that her husband’s family intends it is prospective risk, which is not. The second is the reason the application exists.
These files also defeat the guidance on its own terms. Where the practice is carried out within the family and tolerated by the community, the avenues for recourse an officer is directed to examine are the same avenues the protection stream examines under state protection, and the answer is usually that formal prohibition coexists with non-enforcement. That is a finding about the effectiveness of state protection, made under another name.
Forced and child marriage
The compilation records that women forced into marriage without their consent have been recognised as a social group, on the footing in Vidhani v. Canada (Minister of Citizenship and Immigration), [1995] 3 F.C. 60, that a fundamental right, the right to marry of one’s own choice, had been violated.
Here the identity of fact and factor is total. The engagement, the family’s insistence, the threats on refusal and the age of the intended husband are simultaneously the evidence of persecution and the evidence of hardship. No residue of fact is left once the factors are removed. An officer who excludes the factors has excluded the file.
Honour-based violence
The persecutor is a family member, sometimes several. That makes state protection and relocation determinative in a way unusual even in this area, because the applicant is not fleeing a stranger who may lose interest but a kinship network with a motive to search and a claim to legitimacy in the community.
The guidance provides that an applicant may face hardship in one part of the country but might reasonably be expected to seek relief elsewhere, in which case undue hardship does not exist. For a woman whose family regards her death as a restoration of honour, that is the internal flight alternative inquiry and nothing else. It is also an incomplete version of it. The test settled in Rasaratnam v. Canada (Minister of Employment and Immigration), [1992] 1 F.C. 706 (C.A.), and Thirunavukkarasu v. Canada (Minister of Employment and Immigration), [1994] 1 F.C. 589 (C.A.), requires not only an absence of serious risk in the proposed area but a finding that conditions there are such that it would not be unreasonable, in all the circumstances including those particular to the claimant, to seek refuge there.
That second prong is where a woman’s actual capacity to move is assessed, and the case law shows what it demands. In Syvyryn v. Canada (Minister of Citizenship and Immigration), 2009 FC 1027, Justice Snider set aside a decision because the Division’s second-prong analysis rested solely on the claimant’s twenty years of experience in accounting and contained no analysis of her age, sex or personal situation, as Chairperson’s Guideline 4, Women Refugee Claimants Fearing Gender-Related Persecution, requires, when the documentary evidence showed that women of her age faced significant discrimination seeking employment in Ukraine.
The humanitarian instruction carries the first idea and not the second. It asks whether relocation would eliminate the hardship. It says nothing about whether relocation would be reasonable in this applicant’s circumstances, which is the prong that does the work for a woman with no independent means, no family willing to receive her, and children to move.
Trafficking and domestic servitude
A woman recruited abroad by deception, whose passport was confiscated, who was held in debt bondage and who escaped a network that continues to look for her presents a file in which completed harm and continuing danger are one narrative. Domestic servitude produces the same structure with an employer in place of the network. In both, the past facts are available to the officer and the reason the applicant cannot return is what the same people will do next.
Sexual violence in conflict
Rape used as a weapon, sexual slavery, forced marriage to a combatant and forced pregnancy combine gender, conflict and persecution. These files carry a further difficulty. Where the violence is widespread, the requirement that the applicant be directly and personally affected pushes a decision-maker toward treating generalized suffering as insufficient.
The Court has closed that route in terms. In Diabate v. Canada (Citizenship and Immigration), 2013 FC 129, Justice Gleason held that it is both incorrect and unreasonable, in a humanitarian analysis, to require an applicant to prove that the circumstances she will face are not generally faced by others in her country of origin. That principle has been applied since Kanthasamy, and in Damian v. Canada (Citizenship and Immigration), 2019 FC 1158, Justice McHaffie set aside a refusal because the officer had dismissed evidence of political violence in Colombia not on the basis that it had not been shown to affect the applicant, but because it would affect other Colombians as well.
The pressure toward that error survives the prohibition. Kanthasamy requires the applicant to show she would be personally affected by an adverse country condition, and the line between being personally affected and being differently affected from everyone else is easily lost. It is hardest to hold in this category, where the harm is both individual and endemic. The universality of violence against women becomes, if an officer is not careful, the reason no individual woman qualifies.
The mother who protects her daughter
This configuration deserves separate treatment because it produces two victims in one application. A mother who refuses cutting or marriage for her daughter may face threats from her own family, violence from her husband, exclusion by the community and indifference from the authorities.
The daughter’s position enters through the best interests analysis, which subsection 25(1.3) does not touch. The mother’s position enters through hardship, which it does. The same conduct by the same actors is therefore assessed under two different regimes within one set of reasons, one permitting a full examination of what awaits the child and the other not.
What the profiles share
Four features recur. The persecutor is private rather than the state. The state’s unwillingness or inability to intervene is the pivot of both inquiries. The harm is continuing rather than completed. And the applicant’s own account of what she fled is simultaneously her best evidence of hardship and the material the provision instructs the officer to set aside. A provision that fails across every profile in a category is not encountering hard cases at its margin. It is misdescribing the category.
IV. Children
Kanthasamy holds that the best interests of a child directly affected must be well identified and defined and examined with a great deal of attention, and that children will rarely, if ever, be deserving of any hardship. In Damian v. Canada (Citizenship and Immigration), 2019 FC 1158, Justice McHaffie held that there is no reasonable basis to conclude that a child brought to Canada by her mother acted in deliberate disregard of Canada’s immigration laws, and that holding a parent’s conduct against the child in that way is contrary to a humanitarian approach to the assessment and is unreasonable. Chandidas v. Canada (Citizenship and Immigration), 2013 FC 258, shows what the best interests analysis itself requires. Justice Kane set aside a refusal concerning a nine-year-old being treated for relapsed leukemia, holding that the officer had not identified and defined the child’s interests at all, but had stated the obvious, that she would remain with her parents. The starting point should have been how her follow-up treatment and recovery would best be secured. The officer should then have weighed the concrete consequences of a refusal, which were a return to seek follow-up care in an unknown hospital, with unknown doctors, competing for attention among many other patients where few specialists practise. Three complications arise in this class of file.
The child is frequently a witness and often a direct victim, and exposure to intimate partner violence is itself a harm to the child. That harm is simultaneously a best interests consideration the officer must weigh with great attention and evidence of the risk environment the provision removes from her. Subsection 25(1.3) therefore reaches into an analysis Parliament did not exclude and which the Supreme Court has said requires particular care.
The child’s interests and the mother’s protection are not separable. A child returned with a mother who cannot obtain state protection is a child in an unprotected household, and the assessment depends on a question the officer has been told not to consider.
Where the child is herself the applicant, every element is prospective. A child who fled recruitment by an armed group, or abduction for marriage, or trafficking, is not asserting a completed injury. She is asserting that the recruiters remain. The framework built around a child affected by a decision fits awkwardly, and the provision applies with full force to a person who may have no capacity to construct the required frame at all.
V. The administrative response treats the symptom
Officers may issue a fee-exempt temporary resident permit to out-of-status foreign nationals and their dependants experiencing family violence, conferring short-term status, access to an open work permit or study permit, and health coverage under the Interim Federal Health Program including medical and psychological services. Local offices prioritise processing given the urgency and vulnerability of the client. Between 2019 and 2023 the department issued 594 such permits, approximately 83 percent to women, at a 75 percent approval rate with processing at 44 days as of 2023. A dedicated family violence stream exists within the in-Canada humanitarian category. These instruments save people. They leave the underlying problem where it was.
The permit confers temporary status. It does not resolve the permanent residence application, which is decided under section 25 subject to subsection 25(1.3) by an officer facing the same unperformable separation. A woman who receives a permit in year one and a refusal in year three has been protected from her immediate danger and then returned to it by a provision that prevented the decision-maker from looking at the danger directly.
The guidance confirms the tension from the other direction. Family violence considerations are an enumerated factor, and decision makers are directed to apply Gender-Based Analysis Plus principles. An officer is therefore required to engage with a history of violence while the statute forbids consideration of the factors that history establishes. That is not a demanding standard. It is two standards that cannot both be met.
VI. What follows for counsel
Plead state protection in the guidance’s own vocabulary. Address redress, recourse and the operation of police, courts, human rights tribunals and civil society bodies, rather than framing the submission as unavailable state protection. The analysis is identical. The label determines whether it is received.
Supply the missing reasonableness prong yourself. Where relocation is in issue, do not simply assert continuing danger elsewhere. Set out what the applicant would actually face on arrival in the proposed city: income, housing, childcare, family support, the practical capacity to move. That limb is absent from the instruction, so it has to come from the record.
Anticipate the silence inference. Where the client did not go to the police, explain why in the application rather than on judicial review. In Ndjavera v. Canada (Citizenship and Immigration), 2013 FC 452, Justice Rennie held that the Refugee Protection Division erred in drawing adverse plausibility findings from a claimant’s failure to complain to a police commissioner without properly considering her age, culture, background and experience, being factors set out in the Gender Guidelines. In Hindawi v. Canada (Citizenship and Immigration), 2015 FC 589, the Court held it unreasonable to conclude, without examining the claimant’s particular situation, that her fear was a mere subjective reluctance to seek state protection. Neither decision binds a section 25 officer, and both articulate reasoning she may adopt if given the material.
Use the best interests analysis. It is the least constrained part of the file, and it is underused. Exposure to violence is a harm to the child in its own right and enters through a door the provision does not close.
Plead establishment so it cannot be turned into a sword. In Lauture v. Canada (Citizenship and Immigration), 2015 FC 336, Justice Rennie held that establishment must not be assessed by whether the applicant could carry on similar activities in the country of origin, since on that approach the more successful and civic minded an applicant is, the less likely her application will succeed. In Singh v. Canada (Citizenship and Immigration), 2019 FC 1633, the Court held that an officer cannot use the applicants’ shield against them as a sword.
Chandidas adds a point that belongs in every file of this kind. The officer there examined the family’s establishment in detail, covering work, income, family ties, courses taken, schools attended and community participation, and then found it no more than would be expected after four years, without ever indicating what extraordinary or exceptional establishment would look like. The Court held that this was not reasoning but an informative statement. It also held that the officer should have asked whether a family that integrated into work, school and community throughout the years their daughter was undergoing treatment had achieved more than the expected degree of establishment. Establishment built while surviving something is not ordinary establishment, and the record has to say so. In this class of file, resilience is simultaneously the applicant’s best evidence and her greatest exposure.
VII. What should change
Two remedies are available, and neither requires reading the provision out of the statute. Parliament’s purpose is legitimate and the courts have no warrant to disregard it.
The interpretive remedy is available now and costs nothing. Where the excluded factors and the required hardship rest on the same facts, the mandatory half of the subsection governs. The officer’s task is then to assess hardship fully rather than to police a boundary that cannot be located, and reasons that acknowledge the difficulty and explain how the evidence was used will survive review better than reasons asserting a separation the material does not support.
The legislative remedy is narrower and more obvious. Subsection 25(1.3) could be amended to except applications by a person who has not had a determination under section 96 or subsection 97(1). That preserves the rationale entirely, since avoiding a second determination presupposes a first one, and it removes the provision from the cases where it has no work to do. Parliament has already shown that it knows how to build such a rule with carve-outs: failed claimants may not request humanitarian consideration for twelve months following a final negative decision, subject to exceptions where a child under eighteen would be adversely affected or where the applicant or a dependant suffers a life-threatening medical condition that cannot be treated in the country of origin. Subsection 25(1.3) has neither the precision nor the exceptions.
The woman in the opening paragraph has no conduct to explain away and no credibility problem. Her difficulty is that the only thing she has to say is the thing the officer has been instructed not to hear, unless she says it in different words. That is not a balance between competing objectives. It is a drafting artefact, and it falls on the people least equipped to absorb it.
