It is not often that a case offers an opportunity to clarify the constitutional foundations of the law of judicial review, the ever-present architecture governing review of energy-related decisions. The Supreme Court’s decision in Democracy Watch provided that opportunity. A unanimous Court (per Wagner CJC) held that legislatures cannot oust judicial review on questions of fact or law.^[1] This conclusion is significant not only for the energy sector, but for the very substance of the law of judicial review.

I first review the facts, and then address three issues: (1) the Court’s holding in the case that issues of fact must always be left open for curial review; (2) its conclusion that the reasonableness standard is not constitutionalized; and (3) implications for the energy sector.

Facts

Democracy Watch involves sensational facts. The appellant, Democracy Watch, is a non-profit advocacy organization. It applied to the Federal Court of Appeal for judicial review of a report prepared by the Conflict of Interest and Ethics Commissioner. The report concluded that the then-Prime Minister, Justin Trudeau, had not contravened the Conflict of Interest Act, when he participated in two Cabinet decisions involving WE Charity.^[2] Democracy Watch alleged that, in reaching this conclusion, the Commissioner had committed two errors of law and one error of fact.

The respondent, the Attorney General of Canada, moved to strike the application. The Attorney General advanced two arguments: first, that Democracy Watch lacked standing to bring the application; and second, that the application was barred by s. 66 of the Act. Section 66 provides that the Commissioner’s decisions may be reviewed only on grounds limited to jurisdiction, procedural fairness, or acting or failing to act “by reason of fraud or perjured evidence,” thereby precluding review on the questions of law and fact that Democracy Watch sought to raise.

Before the Supreme Court of Canada, Democracy Watch contended that, properly interpreted, s. 66 does not preclude judicial review at all; and that if it does, the provision is of no force and effect because it is inconsistent with the Constitution Act, 1867, as it would empower the Commissioner to determine the limits of his own jurisdiction and would offend the rule of law.^[3]

Underneath Democracy Watch’s unique facts is the percolating constitutional question: to what extent can legislatures oust the ability of a party to file an application for judicial review in relation to an administrative decision? The answer to this question was facially answered by the Supreme Court’s decision in Crevier.^[4] Crevier held that legislatures must always preserve judicial review over questions of jurisdiction.^[5] But the law of judicial review has developed such that the term “jurisdiction” lacks any real meaning.^[6] This left the scope of the constitutional guarantee of judicial review unclear.

QUESTIONS OF LAW AND FACT

Democracy Watch holds that all aspects of an administrative decision must be open to review, whether the question is one of law, fact, or fairness. A privative clause barring review on any of these issues will be ineffective.

In reaching this conclusion, Wagner CJC rejects the proposition that I advanced, channeling Near JA in Canada (Attorney General) v Best Buy Canada Ltd, that restricting the constitutional guarantee to errors of law meets “any threshold established in Crevier.”^[7] This would include egregious errors of fact that rise to the level of a question of law, like findings based on no evidence, because such errors in the fact-finding process are legal errors.^[8] On this much all parties agreed: as Wagner CJC said in Democracy Watch, “rationality review has a constitutional aspect.”^[9] The disagreement lay in which factual errors should be open to correction, with those in the “legislative intentionalist” camp preserving space for legislatures to bar review on many fact issues.^[10]

But this is not what the Court accepts. Paragraph 64 of the judgment unlocks the case. Wagner CJC says that, after CUPE v New Brunswick Liquor Corp, patently unreasonable findings of fact could not be protected by a privative clause — such a finding would be a jurisdictional error.^[11] Crevier came next, guaranteeing review for jurisdiction, including such errors.^[12] But Crevier did not indicate a departure from what Wagner CJC sees as the basis for CUPE: the idea that an irrational finding could not be protected by a privative clause. Porting that to the current day, the constitutional guarantee tracks a set of principles requiring review, rather than any particular category of error.^[13] An unreasonable error is just as irrational as a patently unreasonable one on this approach. What the Court establishes, then, is that the guarantee tracks a principle rather than a category of this or that error.

Evidence for this comes from Wagner CJC’s own analysis. The authorities Wagner CJC relies on speak only to egregious — patently unreasonable — errors, rather than any error of fact, which is where Democracy Watch ends up. In Toronto (City) Board of Education v OSSTF, District 15, the issue was whether a finding was based on “no evidence.”^[14] In Blanchard v Control Data Canada Ltd, a centrepiece of Wagner CJC’s reasons, Lamer J spoke of “unreasonable findings”, but did so in applying the patently unreasonable standard.^[15]

The distance between these formulations and Democracy Watch is by design. It doesn’t matter so much that only patently unreasonable errors were reviewable. What matters is that a principle of rationality was guaranteed. That principle of rationality now promises review on all facts.

But this headline won’t mean much in day-to-day practice. Courts will continue to review decisions as they have. While some decisions involving facts will now be reviewable, I expect courts to fasten onto a key part of Vavilov: “[t]he reasonableness of a decision may be jeopardized where the decision maker has fundamentally misapprehended or failed to account for the evidence before it.”^[16] In the end, it won’t be easier to make out factual errors once in court.

CONSTITUTIONALIZING REASONABLENESS?

The statute in Democracy Watch is squarely about availability of review, not intensity. Indeed, the Court said that legislative limits on review are ultra vires “not because reasonableness review, per se, is constitutionally entrenched, but because the Constitution guarantees the courts’ role in ensuring that all exercises of public power…are sourced in law.”^[17] However, the issue was not directly in front of the Court.

But Wagner CJC leaves the door open a crack, and the evolution at the heart of Democracy Watch will undoubtedly force debate on the scope of the decision. When the time comes, there is good reason to fully decline the invitation to recognize reasonableness review as a constitutional floor. This was raised in CNR v Alberta Pacific Forest Industries, where the appellant argued in the alternative — before Democracy Watch — that reasonableness review is constitutionalized.^[18]

The starting point is that the guarantee is about availability. Intensity should be a constitutional question at one point only: where a legislated standard is so deferential that it is tantamount to no review at all. In principle, I agree with Khullar CJA, dissenting, in Northback Holdings: the constitutional guarantee and a particular standard of review, such as reasonableness, “are not a package deal logically speaking.”^[19] Vavilov and Democracy Watch both permit legislated standards of review that are consistent with the rule of law—British Columbia has long included, and not changed, the standard of patent unreasonableness in the Administrative Tribunals Act.^[20]

It would be a remarkable assertion of power under ss.96–101 to hold these legislative efforts to prescribe standards unconstitutional because of a constitutionalized reasonableness standard. Three reasons come to mind.

First, Democracy Watch itself counts against a constitutionalized reasonableness standard. Since the constitutional guarantee is about availability, it is not prescriptively tied to reasonableness specifically. Notably, Wagner CJC’s statement of the courts’ role (to ensure administrative action is sourced in law) and his disclaimer about reasonableness are not merely adjacent, but are side-by-side in the same paragraph.

Second, and as a result, for any argument on intensity to be cognizable, it must present the same problem that Democracy Watch deals with — something tantamount to an ouster of review. In other words, legislatures should not be able to do indirectly, via the standard of review, what they cannot do directly by barring review. This principle has considerable vintage as a basic tenet of constitutional law.^[21] We can imagine situations where this might be the case — if review is limited only to bad faith, or some other standard but the test should be deliberately strict to match the concern in Democracy Watch, which was a statute that made review totally unavailable on certain grounds. Something like patent unreasonableness, for example, would pass the test. Though patent unreasonableness is no longer a recognized common law standard of review, it does permit review for legality because it allows for review on all legal issues.

Finally, this strict test is consistent with principles internal to the law of judicial review, and the constitutionalization of reasonableness would run roughshod over them. The power of legislatures to calibrate the relationship between administrative actors and courts — outside of constitutional concerns — is unquestioned. Vavilov endorses it in its articulation of the importance of institutional design choices. That principle is embedded in our law of judicial review. It holds a place in the Court’s holding on rights of appeal, where a highly deferential standard — palpable and overriding error — can be prescribed by the legislature. It generally permits legislatures, as BC has, to prescribe a standard explicitly. To hold such efforts unconstitutional because of a specific common law standard we happen to adopt in 2026 would be an odd retrofit — and it would be inconsistent with Democracy Watch’s principle-based approach. Section 96 has a “judicially nourished luxuriance”, as Laskin once said, but this would be a new level.^[22]

IMPLICATIONS

What are the implications of this decision for the energy sector?

First, the decision clarifies the limits of Parliament’s ability to bar review of decisions that might impact the energy sector going forward. In the interests of speed and efficiency, Parliament might have been tempted to limit review of project approvals or technical assessments, especially on issues of fact and evidence. Democracy Watch will now prevent this. It does so on the theory that “rationality review has a constitutional aspect.”^[23] In other words, it is impossible for Parliament to expressly bar judicial examination of the factual and evidentiary basis for administrative decisions.

Another implication pertains to the existing statute book. Several existing statutes limit or tailor judicial remedies for energy-related decisions. Each of these arrangements will have to be carefully reviewed to determine compliance with Democracy Watch. Consider several arrangements. In Alberta, s.45(1) of the Responsible Energy Development Act provides a right of appeal on questions of jurisdiction or on questions of law, and s.56 of that statute bars all review otherwise.^[24] This arrangement will straightforwardly violate Democracy Watch because it expressly bars factual review. The likely fate of REDA will be a striking down of the ouster to the extent it bars factual review under the right of appeal.

For its part, the Canadian Energy Regulator Act^[25] presents a different, more ambiguous situation. Section 72(1) of that statute permits an appeal from a Commission decision to the Federal Court of Appeal on questions of law and jurisdiction, and s.70 otherwise says that all decisions under the statute are final. The clause at issue in the Canadian Energy Regulator Act is better understood as a “finality” clause rather than an ouster clause. Since judicial review theoretically remains available, this architecture in the Canadian Energy Regulator Act is likely Democracy Watch compliant. However, another part of the statute may be constitutionally suspect: s. 170 of the Act, dealing with decisions of the Pipeline Claims Tribunal, forbids judicial review on issues of fact and law.

Additionally, the Supreme Court’s upcoming appeal in CNR v Alberta Pacific Forest Industries will tell us much about the scope of Democracy Watch.^[26] One issue there is whether a limited right of appeal can effectively oust judicial review under s.18.5 of the Federal Courts Act, which bars judicial review on issues “to the extent that” those issues are “appealable” under a right of appeal.^[27] The Federal Court of Appeal has long said that any appeal — whether to a court or to Cabinet — can effectively bar review on issues not covered by the right of appeal.^[28] Democracy Watch certainly throws doubt on s.18.5: para 73 says that Parliament cannot combine an allocation of jurisdiction to the Federal Courts with a limit on review that would “circumvent the constitutional guarantee of legality.”^[29]

I think the arrangement in CNR channels, rather than directly bars, curial review, leaving it open on issues of fact and law. It permits review on questions of law to the Federal Court of Appeal, while permitting a Cabinet appeal on other issues, including fact and policy, subject to ultimate review in the Federal Courts.^[30] As a matter of statutory interpretation, we cannot look at the Cabinet appeal in isolation, but must look at the entire structural arrangement Parliament put together to channel review: this is the principle of institutional design in practice, one on which Vavilov hung its hat.^[31]

Taking this global look, the arrangement in CNR is constitutionally adequate, unlike the arrangement in Democracy Watch. What disqualified parliamentary oversight there was not that it was weak, but that it was not a court. The Cabinet appeal in CNR ends in curial review. And this read of things falls squarely within language in Democracy Watch: “[p]rovided that courts can exercise their supervisory jurisdiction over the legality of all exercises of public power, the rule of law does not preclude meaningful decision-making by other state actors.”^[32] Because it is constitutionally adequate, the Cabinet appeal becomes one of the remedies a party must exhaust before seeking review. Gleason JA is onto something in saying that s.18.5 merely codifies the idea that parties must exhaust all available remedies.^[33]

CONCLUSION

Democracy Watch will not change, in the short term, how judicial review is conducted. However, it does have implications for legislative power in Canada, and the tools that legislators and regulators can use to streamline or make more efficient regulatory processes. In that sense, the story is not yet over — but Democracy Watch has set the terms of a new debate.

• * Mark P. Mancini is an Assistant Professor at Thompson Rivers University Faculty of Law. He holds a J.D. from the University of New Brunswick and an LL.M. from the University of Chicago Law School, and is a PhD candidate at the Peter A. Allard School of Law at the University of British Columbia.

  ¹ Democracy Watch v Canada (Attorney General), 2026 SCC 28 [Democracy Watch].

• ² Conflict of Interest Act, SC 2006, c 9, s 2, s 66 [Conflict of Interest Act].

• ³ Constitution Act, 1867 (UK), 30 & 31 Vict, c 3, ss 96–101, reprinted in RSC 1985, Appendix II, No 5.

• ⁴ Crevier v AG (Quebec) et al, [1981] 2 SCR 220 [Crevier].

• ⁵ Ibid at 236.

• ⁶ See e.g. Mark Mancini, “Foxes, Henhouses, and the Constitutional Guarantee of Judicial Review: Re-Evaluating Crevier” (2024) 102:2 Can Bar Rev 315.

• ⁷ Canada (Attorney General) v Best Buy Canada Ltd, 2021 FCA 161 at para 60 [Best Buy 2021], citing Crevier, supra note 4; Democracy Watch, supra note 1 at para 71.

• ⁸ See Best Buy Canada Ltd v Canada (Border Services Agency), 2025 FCA 45 at para 11; Schuldt v The Queen, [1985] 2 SCR 592.

• ⁹ Democracy Watch, supra note 1 at para 69.

• ¹⁰ See e.g. Paul Daly, “2023 Developments in Administrative Law Relevant to Energy Law and Regulation” (2024) 12:1 Energy Regulation Q, online: <energyregulationquarterly.ca/regular-features/2023-developments-in-administrative-law-relevant-to-energy-law-and-regulation>.

• ¹¹ CUPE v New Brunswick Liquor Corp, [1979] 2 SCR 227 [CUPE]; Democracy Watch, supra note 1 at para 64.

• ¹² Crevier, supra note 4.

• ¹³ Paul Daly, “The Supreme Court of Canada’s Decision in the Democracy Watch Case: Democracy Watch v Canada (Attorney General), 2026 SCC 28” (30 July 2026), online: Administrative Law Matters <administrativelawmatters.com/blog/2026/07/30/the-supreme-court-of-canadas-decision-in-the-democracy-watch-case-democracy-watch-v-canada-attorney-general-2026-scc-28>.

• ¹⁴ Toronto (City) Board of Education v OSSTF, District 15, [1997] 1 SCR 487.

• ¹⁵ Blanchard v Control Data Canada Ltd, [1984] 2 SCR 476 at 499–500.

• ¹⁶ Canada (Minister of Citizenship and Immigration) v Vavilov, 2019 SCC 65 at para 126 [Vavilov].

• ¹⁷ Democracy Watch, supra note 1 at para 71.

• ¹⁸ Canadian National Railway Co v Alberta Pacific Forest Industries Inc, SCC Docket No 42092 (leave to appeal granted), online: Supreme Court of Canada <scc-csc.ca/cases-dossiers/search-recherche/42092> [CNR].

• ¹⁹ Northback Holdings Corporation v Alberta Energy Regulator, 2025 ABCA 186 at para 228, Khullar CJA, dissenting [Northback Holdings].

• ²⁰ Administrative Tribunals Act, SBC 2004, c 45, s 58.

• ²¹ See e.g. Amax Potash Ltd v The Government of Saskatchewan, [1977] 2 SCR 576 at 591; Dominion Stores Ltd v R, [1980] 1 SCR 844 at 858.

• ²² Bora Laskin, “Municipal Tax Assessment and Section 96 of the British North America Act: The Olympia Bowling Alleys Case” (1955) 33:9 Can Bar Rev 993.

• ²³ Democracy Watch, supra note 1 at para 69.

• ²⁴ Responsible Energy Development Act, SA 2012, c R-17.3, ss 45(1), 56 [REDA]; Northback Holdings, supra note 19.

• ²⁵ Canadian Energy Regulator Act, SC 2019, c 28, s 10 [CERA].

• ²⁶ CNR, supra note 18.

• ²⁷ Federal Courts Act, RSC 1985, c F-7, s 18.5.

• ²⁸ Canadian National Railway Company v Scott, 2018 FCA 148.

• ²⁹ Democracy Watch, supra note 1 at para 73.

• ³⁰ See Canadian National Railway Company v Emerson Milling Inc, 2017 FCA 79, Stratas JA.

• ³¹ Vavilov, supra note 16.

• ³² Democracy Watch, supra note 1 at para 69.

• ³³ See CNR v Alberta Pacific Forest Industries, 2025 FCA 160 at para 26.