Dear readers,
As summer approaches, I wanted to advise that there will be a few days this summer where there will be no SEAR. These dates include the following:
July 5
July 12
August 2
August 23
Thank you for your understanding!
Northback Holdings Corporation v Alberta Energy Regulator, 2025 ABCA 186 (May 27, 2025)
Category: Constitutional minimum of judicial review.
*Small updates and editorial adjustments made post-publication, June 1*
Context: This is an appeal of a decision of the ABKB, dismissing applications for judicial review (I reviewed the ABKB decision in Issue #119). In the underlying decision, the Alberta Energy Regulator decided not to approve Benga Mining’s (now Northback Holdings Corporation) Grassy Mountain Coal Project. Section 45(1) of the Responsible Energy Development Act [REDA] provides an appeal (with permission) to the Court of Appeal “on a question of law or jurisdiction.” Section 56 entrenches a broad privative clause that appears to preclude review over all other questions. The constitutional issue is whether arguments based on fact and mixed fact and law are precluded by the combination of ss.45 and 56.
Issue and Holding: The majority (per Fagnan and Shaner JJA) conclude that the constitutional issue is not properly before the Court: the appellants failed to provide timely notice of the constitutional issue, and they also failed to raise the constitutional issue in its applications for permission to appeal under s.45.
Khullar CJ dissented, addressing the constitutional head-on. She would have concluded that the privative clause in s.56 is unconstitutionally capable of barring review, because “the constitutional minimum includes review on questions of fact and law” [64].
Analysis: This is a significant case that whets the appetite for the Supreme Court’s decision in Democracy Watch, on appeal from Democracy Watch, 2024 FCA 158 (Issue #154). In particular, Khullar CJ offers a compelling set of reasons articulating her framing of the constitutional guarantee of review. In so doing, she spends quite a bit of time engaging my work on the topic, which takes a different position from her: see here. Readers can certainly make up their own minds. For my part, I think the majority properly approached the issues, and I ultimately disagree with Khullar CJ’s perspective on the constitutional issue.
First, the majority’s analysis. Besides dismissing the appeal on the basis of the notice/argument issue, it also ventured a proposal for how to solve the issue in this case. Simply put, the majority concludes there was no need for a detailed discussion of the constitutional guarantee: the starting point should be the language of s.45 itself [53]. Recall that s.45 of the REDA provides for permission to appeal on a question of law or jurisdiction. Whatever the scope of the constitutional guarantee, language covering off “law and jurisdiction” would satisfy any plausible definition of the guarantee, which would encompass strictly legal questions and the tricky category of factual questions falling under the label of “jurisdiction.” So, s.45 can be seen as channeling curial review, obviating the need to re-define its scope in the abstract.
The question is one of simple statutory interpretation. As the majority states, “[n]o rationale has been put forward as to why a satisfactory solution to the appellants’ constitutional concerns would not reside in this Court’s interpretation of s 45” [43]. Given the prudential problems in this case—arguable lack of notice, lack of findings on the scope of the permission right of appeal—wide-ranging constitutional analysis would be inappropriate when s.45 already channels curial review.
In this sense, the majority’s approach mirrors the Federal Court of Appeal’s approach in Emerson Milling, 2017 FCA 79. There, the Court dealt with an appeal from a Canadian Transportation Agency decision, with the appeal limited to questions of law or jurisdiction. Noting the difficulty of defining the term “jurisdiction” in the current state of Canadian administrative law, Stratas JA nonetheless noted that the original meaning of the term “jurisdiction” in the statute was largely the same as questions of law, but also included factually-suffused procedural questions (Emerson Milling, 18-19). Stratas JA began with the words of the statute and their original meanings, not with a broad constitutional analysis.
So, the resolution of this case was not as complex as it might seem. Nonetheless, Khullar CJ addresses the constitutional problem directly, concluding that notice was not required [124]. Hers is a dissenting opinion, but given its depth and strength—and given the stakes in Democracy Watch—it is worth unpacking (note, though, that the institutional design of the REDA is quite different from the design in Democracy Watch).
Khullar CJ’s opinion can be distilled into four general propositions:
The constitutional guarantee of judicial review must be re-formulated away from the concept of jurisdictional error, given the difficulty of defining the concept [186]. On this, she explicitly distinguishes herself from the view expressed by de Montigny CJ in Democracy Watch, at para 75 [185].
One possible reformulation is to “reconstruct the constitutional minimum as curial review on questions of law” [198]. She notes that I, alongside some judicial opinions, advance this perspective [199]. The perspective is this: judicial review is fundamentally about keeping decision-makers within legal limits, primarily statutory boundaries. The theory of jurisdictional error is premised on the idea that decision-makers should not be the final arbiters of their statutory boundaries: this would, in effect, constitute them as a s.96 court. Crevier held that legislatures cannot, via the delegation of power, create shadow s.96 courts. A constitutional guarantee limited to questions of law (with allowance for factual errors rising to the level of a question of law) adequately captures the central features of the constitutional guarantee under Crevier. This perspective has the benefit of coherence with this foundational principle of administrative law, while avoiding the need for law office metaphysics about the precise definition of “jurisdiction” post-Vavilov.
Khullar CJ advances three arguments against the question of law formulation:
Addressing my paper, she argues that the question of law formulation “rests on unproven assumptions about the behaviour of administrative decision makers” in that they may have a temptation to exercise power beyond their authority [218-219].
The certainty benefits of the question of law reformulation are “precarious” [220-221]. That is because limiting the constitutional guarantee to questions of law will incentivize litigants to plead that certain questions of fact rise to the level of a question of law, leading to “artful pleading” [221].
The constitutional guarantee, as a matter of principle and precedent, is not and should not be reserved to questions of law. For one, administrative law is not just about compliance with statutory boundaries, since there are also common law constraints on administrative power [218]. What’s more, a failure to recognize this can insulate administrative power from appropriate review: “[a]t the extreme, arbitrary or careless decisions affecting important interests would be free from judicial oversight as long as the arbitrariness or carelessness “only” relates to the interpretation or weighing of evidence, the finding of facts, or the application of the law to the facts” [222].
Her formulation, then, is to preserve judicial review over questions of law and fact. However, she does not believe that reasonableness review is the constitutional minimum of judicial review, differing in this respect from the FCA in Best Buy, 2021 FCA 161 (see Issue #4), which also relies on Professor Daly’s work (see para 118 of Best Buy).
Each of these points are worthy of response:
As far as it goes, Khullar CJ is right that the question of law formulation is motivated by a concern for the arrogation of power beyond statutory limits. But if this motivation is an “assumption,” it is one rooted in the history of the law of judicial review and our own positive law. S.A. De Smith put it best in an early edition of his important textbook on English administrative law: if decision-makers had final curial review over their statutory limits, it would offer “temptations that not all would be able to resist.” Khullar CJ cites my work for the temptation idea, but it is not an original thought on my part. That is clear on a review of some of the key cases. Why does Laskin CJC, in Crevier, worry about the creation of shadow s.96 courts? It is because s.96 courts have inherent jurisdiction, with power to opine on the limits of their authority subject to appeal. If legislatures could create shadow s.96 courts in this manner, it would create islands of unreviewable power. The normative justification behind s.96—true, a mere appointing provision—reinforces De Smith’s point.
It is fair for Khullar CJ to worry about artful pleading, but there are good reasons to think the question of law formulation is better than the alternative on this front. For one, while parties do attempt to bootstrap factual issues into rights of appeal limited to law, this is not a new problem. Courts have the tools to navigate the boundary between law and fact. For example, in Best Buy, 2025 FCA 45, Stratas JA compiled the authorities that stipulate which egregious errors of fact will rise to a question of law (see para 11). This includes failure of reasoning leading to a legal error (see, for an example of this, Alexion Pharmaceuticals, 2021 FCA 157, Issue #3) and failure to abide by legal principles governing fact-finding. However, not all factual errors are included, which maps onto the previous articulation of jurisdictional error, in which not all factual errors were reviewable.
Khullar CJ’s worry about arbitrary factual findings is arguably covered by Stratas JA’s framing in Best Buy 2025. As Stratas JA says in Best Buy 2025, “[j]ust about anything that can be raised in a separate application for judicial review can be raised in a statutory appeal where only ‘questions of law’ can be raised” [11]. This would include at least some of the examples Khullar CJ lists of factual errors that would, in her view, be excluded from a question of law formulation. For example, her worry about “arbitrary” decision-making is unlikely to be a concern under a right of appeal on questions of law, since “arbitrary” findings of fact (ie) “no evidence” is a well-established legal error (see Schuldt v The Queen, [1985] 2 SCR 592).
I worry that Khullar CJ’s formulation does not include meaningful limits on a court’s ability to review for factual/evidence errors more broadly, although in the end we are really debating a very narrow band of factual and evidential issues. Nonetheless, there is a point of principle. It is sometimes argued that judicial review is about upholding the rights of individuals in the abstract, which might suggest that the constitutional guarantee should also pay close attention to flaws in administrative action wherever they occur. Upholding the perspective of the individual is certainly a function of the law of judicial review. But defining the function according to this theory alone risks detaching judicial review from its legal limits, while missing that the law of judicial review does not run, as a constitutional matter, to all manner and form of state action. To the extent this is incoherent, it has arguably always been the case: there have always been “gaps” in the supervisory jurisdiction of superior courts.
Accordingly, Khullar CJ’s view risks missing that, as another judge of the ABCA has noted, judicial review exists “…within legal constraints” (see McCargar v Metis Settlements General Council, 2025 ABCA 33, Issue #166). On one hand, she properly notes that “[a] reconstruction that respects the precedents cannot exclude review on questions of fact (and mixed fact and law)” [215]. This would include the law around proper fact-finding, and factual errors parasitic on a legal error. As noted above, the question of law formulation properly encompasses these errors, and so Khullar CJ’s criticism may miss the mark. But on the other hand, any reconstruction “that respects the precedents” cannot allow for an unbounded review for factual and evidentiary matters, because any reviewable factual matters must have a close tie with the law. When Khullar CJ summarizes her framing of the constitutional guarantee at various points, it does seem unbounded in this manner: “I conclude that the constitutional minimum includes review on questions of fact and mixed fact and law” [243]; “[the constitutional guarantee] must include review on questions of fact and mixed fact and law” [225]. But this, itself, is not the rule that emerges from the precedents. It is true that patently unreasonable findings of fact gave rise to a jurisdictional error under the previous law. But these were patently unreasonable findings. The cases do not disclose an unbounded power of judicial review over all questions at all times. I fear Khullar CJ’s reasons can be read in this manner, decoupling judicial review from the enforcement of legal limits, although—again—the practical difference may be exceedingly minimal.
A point of agreement with Khullar CJ: I also do not believe that the constitutional guarantee can be defined according to reasonableness. The question raised by the constitutional guarantee is the category of questions that are reviewable in the first place. While the law of judicial review has moved from a “scope of review” framework (one where deference is manifested in the types of questions that are beyond review) to an intensity of review framework (where the general question is how, not whether, courts conduct review), legislatures often speak in “scope of review” language. This is Khullar CJ’s point. She notes that “My conclusion about the content of constitutional minimum does not depend on accepting that reasonableness review is constitutionally entrenched. They are not a package deal logically speaking” [228]. This is true: the constitutional guarantee is generally about whether something can be reviewed, not how it is reviewed, so reasonableness and the constitutional guarantee are not a “package deal.”
There are other aspects of Khullar CJ’s opinion worth discussing, but this is enough for now. As de Montigny CJ said in Democracy Watch, “this most complex and vexed question” will need to be resolved by the SCC (Democracy Watch, at para 78). That this question is difficult is evident: not only do courts differ among themselves, but judges on the same court do not see eye to eye on the question. There are multiple options on the table for the SCC.
Disclaimer: Nothing in the SEAR is legal advice. The SEAR is not designed to be comprehensive. It is designed to collect some cases that I believe are interesting to practitioners and students, and to provide my quick thoughts. Not all of the cases presented are fully analyzed ie) I’ve selected the most relevant issues in some of the cases. Please complement the SEAR with your own research and study. Any mistakes are my own