Introduction
The best interests of the child appears throughout Canadian law, but it does not do the same job everywhere. In family law it decides cases: the child’s welfare is the governing rule, and the outcome follows from it. In administrative law it does something narrower. It constrains discretion rather than directing the result. The State may act against a child’s interests, but it must account for that choice, and it is the accounting, not the power to override, that the law polices. The principle is rooted in compassion. It is governed by justification.
Humanitarian and compassionate relief under section 25(1) of the Immigration and Refugee Protection Act shows this structure at its clearest. An officer deciding an H&C application weighs establishment, inadmissibility, and public policy, and must also consider the circumstances of any child directly affected. Baker and Kanthasamy require that the child’s interests be engaged in substance. Yet judicial review after Vavilov keeps turning up the same defect: the factor is named, and then it does not work. The reasons acknowledge the child and reach a conclusion, but nothing connects the two. Courts intervene not because the result is harsh, but because the route to it is missing.
This is why the best interests of the child operates as a standard of justification rather than a promise of relief. A decision can go against the child and still be reasonable, but only if the reasons show how the child’s circumstances were actually weighed, not merely that they were noticed. Immigration law makes the wider point visible: where vulnerability is a mandatory consideration, legality turns on explanation, not on outcome.
The Development of the Best Interests of the Child in Canadian Law
The modern principle begins with the 1989 United Nations Convention on the Rights of the Child, which Canada ratified in 1991. Article 3 requires that the interests of the child be a primary consideration in all actions concerning them: a primary consideration, not the paramount one, a distinction that will matter. The Convention is more than an aspiration. It sets an interpretive norm, requiring a decision-maker to ask what a given exercise of State power does to a child’s development, protection, and participation. In Baker v Canada (Minister of Citizenship and Immigration), [1999] 2 SCR 817, the Supreme Court used that framework to fold best-interests obligations into administrative discretion, before any statute told it to.
Domestically, the principle matured first in family law, where it holds paramount status. The Divorce Act and provincial custody and protection statutes direct courts to decide by the welfare of the child: emotional stability, continuity of care, and, as the child matures, the child’s own views. Immigration law borrowed the child-centred method but not the paramountcy. Baker required officers to account for children’s interests in H&C applications even before the Immigration and Refugee Protection Act codified the duty, and Kanthasamy v Canada (Citizenship and Immigration), 2015 SCC 61, confirmed that the account must be real: a substantive assessment, not a recital.
Three features define the principle across these settings. It carries primacy where a decision is directly about a child. It is flexible and contextual rather than fixed. And it demands reasoning from the child’s point of view, built on evidence of the particular child’s circumstances and vulnerabilities. Under Canada (Minister of Citizenship and Immigration) v Vavilov, 2019 SCC 65, that reasoning must be transparent and intelligible. But flexibility cuts both ways. The same openness that lets the principle adapt to each child also lets a decision-maker nod to it and move on, subordinating the child’s interests to inadmissibility or enforcement instead of assessing those concerns through them.
Administrative Law and the Duty to Consider the Best Interests of the Child
Vavilov reshaped the frame around all of this. A decision must now be justified, transparent, and intelligible against its legal and factual backdrop. In an H&C application, that means an officer has to engage the substance of the record, not restate the language of the section. Procedural fairness rises with the stakes in the same way. When removal reaches children, the consequences edge toward the territory section 7 of the Canadian Charter of Rights and Freedoms protects, and the expectation of a genuine confrontation with the evidence rises accordingly.
Under that standard, the doctrine of relevant factors does real work. Some considerations are not optional, and a decision that ignores one is unreasonable for that reason alone. The best interests of the child is such a consideration. Its weight stays contextual, and an officer may give it little in a given case, but the reasons have to show how it was assessed and set against the objectives of immigration control. The Federal Court keeps quashing refusals where a child’s circumstances are acknowledged but never actually assessed, or where the child’s evidence is brushed aside without a word of explanation.
The difficulty is structural. Officers decide files in volume, under statutory language broad enough to leave them wide latitude, and deference respects that role. Vavilov withdraws it only where the reasoning fails to answer the constraints the statute, the record, and the submissions impose. Courts accept that H&C decisions call for polycentric balancing. They do not accept a conclusion that reads as predetermined, or one that floats free of the evidence.
The dividing line in the cases is consistent, and it is not whether the factor was mentioned. It is whether the factor shaped the reasoning. That means tying the evidence to concrete consequences for the particular child: an interrupted education, a course of medical treatment that ends at the border, a relationship severed by removal. The interests are not paramount, as they would be in a custody dispute, but they can move the outcome, and they cannot be filed behind a parent’s conduct as though they were secondary to it. Courts step in where the analysis dissolves instead into generic hardship language or speculation about conditions abroad, and they ask for something specific: an explanation of why the child’s interests were outweighed, not an assurance that they were considered.
All of this sits at the meeting point of two forces in administrative law. Discretion opens room for flexibility in exceptional cases; the demand for justification disciplines how that room is used. The best interests of the child lives exactly there. It does not end the balancing. It requires the State to show its work. Where the reasons do not, the decision stops being defensible in Vavilov’s sense, and it goes back.
Immigration Law as a Lens
Section 25(1) lets the Minister grant permanent residence or an exemption on humanitarian and compassionate grounds, and it expressly requires that a directly affected child’s best interests be considered. Operational guidance tells officers to run a child-focused analysis and to connect the evidence to concrete consequences. After Kanthasamy that analysis has to look forward, not lean on speculation about how the family might manage abroad, and it has to run through the whole assessment rather than sit in a paragraph of its own. The scheme still lets other considerations count, inadmissibility and public safety among them, which keeps it flexible and, at the same time, sets up the tension the cases turn on.
Judicial review shows what that tension produces. Courts have set aside decision after decision where the child’s interests were acknowledged and then left unassessed: where an officer ignored medical evidence, discounted a child’s integration into school, or fell back on generic hardship reasoning. They have upheld decisions, too, where the reasons explained plainly why a child’s circumstances, real as they were, did not outweigh the other objectives in play. The two lines of cases differ less in result than in justification.
So, the case law treats the best interests of the child as structurally influential rather than outcome-determinative. It guarantees no relief. It guarantees a transparent explanation of how the factor shaped the reasoning, and the failures that draw intervention are failures of omission, speculation, or formula, not disagreements with the final call.
Procedural fairness usually travels with these errors. An officer who doubts evidence about a child without giving notice, or who lets a parent’s misconduct eclipse the child’s situation without analysing it, weakens the decision on fairness grounds as well. Volume pressures push toward shortened reasons. The courts have not relaxed the requirement of an individualized assessment to meet them.
Most of these cases end the same way, remitted for redetermination. The cycle repeats, and it carries a plain message: efficiency is not a substitute for explanation. Inside the H&C framework, immigration law becomes a working demonstration of how a discretionary system has to answer for vulnerability.
The Best Interests of the Child as a Measure of Justification
The pattern is easiest to see in the Federal Court’s recent decisions, where the best interests of the child becomes the point at which the reasons either hold together or come apart.
In Rubio v Canada (Citizenship and Immigration), 2025 FC 609, Justice Diner set aside an H&C refusal because the officer broke the analysis into segments and turned the applicant’s adaptability and establishment into reasons to discount hardship rather than counting them, as they should have been, as points in her favour. The updated evidence went unaddressed; the holistic assessment Kanthasamy demands never happened. The best interests analysis failed as part of that larger collapse, because the reasons never showed how the child’s circumstances were folded into the whole.
Acknowledgement without content recurs in Chaudhray v Canada (Citizenship and Immigration), 2025 FC 1278. The officer named the best interests of six Canadian-born children but never said what those interests actually were, and built the analysis on a critical factual error about how familiar the children were with Pakistan, the finding of familiarity resting on brief trips taken when the children were very young. Removing children from their country of citizenship carries serious consequences, and Vavilov asks for justification proportional to them. These reasons did not supply it.
The shortcut in Dedvukaj v Canada (Citizenship and Immigration), 2024 FC 1300, was adaptability. Rather than identify what the children needed, the officer assumed they could adjust to relocation and reasoned from there, trading the mandated child-centred inquiry for a hardship-minimization exercise. Justice Ahmed described the analysis as an appraisal of what an eight-year-old could adapt to and the hardship she could endure, and called it the antithesis of the analysis the law requires. Asking whether a child can cope is not the same as asking what the child’s interests require, and the substitution made the decision unreasonable.
Nagra v Canada (Citizenship and Immigration), 2025 FC 1942, shows the hardship-threshold error in its plainest form. The officer accepted that the applicant, a grandmother, and her grandchildren were closely bonded, then gave the bond minimal positive weight on the ground that the children would suffer no significant harm, and leaned on speculation about who might care for them instead. The Court held that this applied a threshold test for hardship rather than a determination of best interests. It is one thing to treat the factor as non-determinative; it is another to acknowledge the children’s benefit from the relationship and then assign it almost no weight. That is the wrong framework, and the decision fell for using it.
And where the reasoning is merely perfunctory, the outcome is the same. In Tuyebekova v Canada (Citizenship and Immigration), 2022 FC 1677, the officer noted the possible impacts of removal but retreated into generalities about parental caregiving and contact from a distance, without engaging the evidence that mattered most: one child’s special needs, the grandmother’s role, the supporting medical record. Thin reasoning on the children’s interests, joined to a broader failure to grapple with the central hardship evidence, sent the decision back for redetermination.
Read together, the cases show the best interests of the child doing a particular job: it is the place where a court tests whether the reasons cohere. A decision can refuse relief and survive review, provided the analysis explains how the child’s circumstances were weighed against the competing statutory objectives. A decision that recites the factor without showing its influence will fail, even when the same conclusion was open to the officer on a properly reasoned path.
This is the structure of justification Vavilov describes. Discretion survives, but only where the reasoning discloses a rational line from the evidence through the statutory framework to the result. The best interests of the child works on both sides of that line at once: a substantive consideration the officer must assess, and a methodological check on whether the discretion was exercised lawfully. It asks for an assessment grounded in the child’s actual life, and it signals whether the officer stayed within legal bounds.
None of this makes H&C relief a child-paramount regime. It makes the competing considerations answerable to the child’s situation. Public policy, inadmissibility, and enforcement can still prevail, but only after the reasons explain why they outweigh the concrete consequences for the child. Absent that explanation, the decision reads as predetermined, and it is no longer defensible. Immigration law, seen this way, is a practical demonstration of a wider administrative principle. A mandatory consideration shapes the reasoning process, not the outcome: it does not dictate the result, but it sets the quality of justification needed to reach one. The best interests of the child is, in the end, a measure of legality. A decision is reasonable not because it lands on a compassionate result, but because it shows, transparently and intelligibly, how compassion was weighed in the exercise of State power.
The Applicant’s Burden and the Limits of the Duty
The cases to this point tell one side of the story. There is another, and an honest account has to reckon with it, because the Federal Court of Appeal has been clear that the duty to analyse a child’s interests is neither unbounded nor uniform.
Its shape depends first on what the applicant put in issue. In Owusu v Canada (Citizenship and Immigration), 2004 FCA 38, the Court held that the obligation to consider the best interests of a child arises only when the application makes sufficiently clear that it relies on that factor, and that the applicant carries the burden of adducing the evidence. Mr. Owusu’s submissions raised his children in a single sentence of a seven-page letter. The Court held that applicants who leave pertinent information out of their written material do so, in its words, “at their peril,” and it declined to fault the officer for failing to develop what the applicant had barely raised. An officer is under no duty to interview, to hunt for a stronger case than the one filed, or to shore up a submission the applicant left thin. Where the record is sparse, a sparse analysis can be reasonable.
Its shape depends, too, on what the reasons must actually say. In Hawthorne v Canada (Minister of Citizenship and Immigration), 2002 FCA 475, the Court held that an officer’s assessment proceeds on a premise that need not be stated in the reasons at all: that, absent exceptional circumstances, a child’s interests favour the parent remaining in Canada. The officer’s real task is the practical comparison of benefit and hardship, the two being, in the Court’s phrase, “two sides of the same coin.” Reasons that leave the obvious unstated are not deficient for that reason.
And its weight is not fixed. Legault v Canada (Minister of Citizenship and Immigration), 2002 FCA 125, confirms that the best interests of the child, once identified and weighed, can be outweighed by other considerations. An officer who gives the factor genuine attention and still refuses relief has not erred merely by refusing. The child’s interests are a primary consideration, not a trump.
Set against the refusals quashed above, this can look like a contradiction. It is not. The two lines resolve into a single and more precise rule, and the resolution is the point. What the recent decisions require is not maximal articulation in every file. It is articulation proportional to what the applicant placed in issue. Rubio, Chaudhray, Dedvukaj, Nagra, and Tuyebekova are not cases of officers declining to spell out the obvious or answering brief submissions briefly. They are cases where the applicant squarely raised the child’s circumstances, with evidence, and the officer responded with a hardship threshold, a factual error, a generic assumption, or silence. Owusu and Hawthorne mark the floor of the duty; Kanthasamy and the decisions applying it govern what happens once an applicant has built above it.
Read this way, the counter-line does not weaken the standard of justification. It calibrates it. The obligation to show one’s work is triggered and measured by the case the applicant actually makes, which is why the same terse analysis can be reasonable on one record and fatal on another. The variable is not judicial mood. It is the state of the record and the quality of the engagement it called for.
Reasoning with the Child in View
What does adequate reasoning actually look like? It starts by naming the children affected, setting out the evidence that bears on them, and stating what removal would do to their circumstances. School records, medical reports, and personal statements get addressed on their terms, not gestured at in the abstract.
The analysis moves in sequence. The officer identifies the child’s situation, works through the statutory factors, traces the consequences of removal, and only then weighs those consequences against the countervailing considerations: inadmissibility, immigration history, and the rest. A general assurance that the child will adapt, or that relocation is manageable, does not meet the standard when specific evidence points to specific risks.
Where more than one child is involved, each is assessed on their own facts. Age, health, and degree of establishment produce different impacts, and the reasons have to reconcile them rather than resolve them into a single sentence. A collective conclusion with no individual reasoning is the formula the courts keep rejecting.
Adequate reasons engage the record. The passages that matter in a report or a letter are addressed directly, with an explanation of how they bore on the decision; where the officer disagrees with the evidence, the reasons say why. Silence in the face of contradictory material is what so often tips a decision into unreasonableness.
Fairness enters wherever credibility does. If the reliability of a document about a child is in doubt, the affected party is entitled to notice and a chance to respond before the officer draws an adverse conclusion.
Around the individual file, ordinary institutional habits help internal review, training, and record-keeping that keep the analysis tethered to a visible line from evidence to finding to outcome. The standard is always justification, not result. An officer may give the best interests of the child limited weight, but only when the reasons make transparent how that weight was reached.
Conclusion
The line from Baker through Kanthasamy is settled on one point: the best interests of the child must be assessed in substance and tied to the evidentiary record. Vavilov review keeps exposing the same shortfalls, whether reasoning that collapses into formula or reasoning that never confronts what removal would do to the child.
The statute asks for child-specific reasoning, and the Federal Court keeps setting aside refusals that acknowledge the factor without weighing it. The problem is never the outcome as such. It is the missing explanation of how the child’s circumstances bore on the result.
These cases sit within immigration law, but the principle they enforce is not obviously confined to it. Discretion is wide across administrative law, and it is routinely held in check by considerations a statute makes mandatory. Where such a factor also engages a vulnerable person, the same logic would seem to apply: the reasons have to engage the record and explain why the competing considerations prevail, and a decision that cannot do that reads as detached from its own facts. Whether that broader reading holds is not something the immigration cases can settle on their own. It would need testing against mandatory-vulnerability factors in other discretionary schemes, and this article does not undertake that work.
What the immigration jurisprudence does establish is narrower and firmer. Within H&C decision-making, the best interests of the child operates as an interpretive constraint. It demands attention in proportion to the impact on the child and a reckoning with evidence that cuts the other way. Courts that intervene are not substituting their own view of the merits; they are refusing to accept reasons that never trace a coherent path from evidence to outcome. The steady stream of remittals suggests a doctrine still consolidating rather than one fully settled, and its reach beyond immigration remains to be worked out. What is already clear is the core requirement: where the factor applies, the decision-maker has to show how the child’s interests were considered, and why the balance came out where it did.
