Judicial Review on the Vires of Subordinate Legislation: Full Vavilov, Partial Vavilov or No Vavilov?

By: Shaun Fluker

Cases Commented On: Auer v Auer, 2022 ABCA 375 (CanLII) and TransAlta Generation Partnership v Alberta (Minister of Municipal Affairs), 2022 ABCA 381 (CanLII)

PDF Version: Judicial Review on the Vires of Subordinate Legislation: Full Vavilov, Partial Vavilov or No Vavilov?

This comment examines two decisions issued concurrently by the Alberta Court of Appeal in late November 2022 that reject the application of a standard of review analysis when reviewing the vires (aka legality) of a ‘true’ regulation, (the need for the modifier is explained below). This is a topic that I have casually followed for some time. In 2016 I wrote Does the Standard of Review Analysis Apply to a Vires Determination of Subordinate Legislation? and in 2018 I wrote Judicial Review on the Vires of Subordinate Legislation. Together these earlier posts describe an uncertainty that has reigned for years over whether a standard of review analysis applies to the vires determination of subordinate legislation. In its overhaul on standard of review in Canada (Minister of Citizenship and Immigration) v Vavilov, 2019 SCC 65 (CanLII) (Vavilov), the Supreme Court of Canada did not explicitly address this question (for my overview on standard of review under Vavilov see Vavilov on Standard of Review in Canadian Administrative Law). The uncertainty has evolved into a jurisprudential conflict. In Portnov v Canada (Attorney General), 2021 FCA 171 (CanLII) (Portnov), the Federal Court of Appeal ruled that a Vavilov standard of review analysis applies to the vires determination of regulations (Portnov at paras 23 – 28; see more recently Innovative Medicines Canada v Canada (Attorney General), 2022 FCA 210 (CanLII)). In Auer v Auer, 2022 ABCA 375 (CanLII) (Auer) and TransAlta Generation Partnership v Alberta (Minister of Municipal Affairs), 2022 ABCA 381 (CanLII) (TransAlta Generation) the Court of Appeal rules that Vavilov may partially apply to some regulations but not ‘true’ regulations (Justice Feehan departs from the majority in Auer on this point: Auer at para 117)).

Subordinate Legislation: Complicated and Non-Transparent

Before diving into the confusion, I begin with a few basics on the structure of our government. The authority to legislate is inherent to the legislative branch, however a legislature may delegate this authority, and it is delegated to the executive branch in almost every statutory framework. Legislation enacted by the executive branch is referred to as ‘subordinate’ or ‘delegated’ legislation because it is made under the authority of a statute passed by the legislative branch. The key point here is that authority to enact subordinate legislation must ultimately be sourced in a statute. As it turns out, this type of delegation is very popular, and most of the legislation enacted in Canada these days takes the form of subordinate legislation. There are several reasons for this, including that the technical and detailed content of modern regulatory programs simply cannot be legislated without the expertise of the executive and public service, and that the need for specific timing on enactments and amendments can make the legislative process in the elected assembly unworkable. Governments now tend to rely heavily on subordinate legislation, so much so that a democratic deficit has been created because of the amount of legislative activity that occurs outside of the elected assembly. In extreme instances, a legislature will enact a skeletal statute which effectively delegates all actual rule-making to the executive, which is problematic for democracy because delegated lawmaking is usually not an open and transparent process.

Subordinate legislation has the same the force of law as a statute, however the process by which each is enacted has crucially important distinctions. A statute has its beginnings as a bill tabled in the legislative assembly. The bill must pass through the legislative process, which in Canada includes three readings in the elected assembly before it can become law. A bill becomes a statute after it passes third reading and receives Royal Assent. All public statutes enacted by the Alberta legislature are published by the King’s Printer in accordance with the King’s Printer Act, RSA 2000, c K-2. In contrast, subordinate legislation does not pass through the legislative process. Very often, subordinate legislation is enacted by executive order without meaningful public notice and no opportunity for public comment. Municipal bylaw-making would be a common exception to this democratic gap; for example, the Municipal Government Act, RSA 2000, c M-26 stipulates an open legislative process as well as other public participation measures on lawmaking by municipalities in Alberta. Some jurisdictions also have committees with elected members of the legislature who periodically review subordinate legislation after it is made by the executive branch, but even in these instances, there is no point-in-time scrutiny by the elected assembly of the enactment of subordinate legislation. Too often, delegated lawmaking closely resembles legislating in secrecy.

Regulations are just one type of subordinate legislation. Other common iterations include bylaws enacted by municipalities; rules, instruments, and directives enacted by administrative agencies; guidelines, codes, orders, and protocols enacted by any person with delegated legislative powers. And while regulations or orders are usually the form of subordinate legislation enacted by the political executive – cabinet or an individual minister – regulations or orders can also be enacted by other entities within the executive branch, such as administrative agencies. It all depends on what the governing statute allows for. The Securities Act, RSA 2000, c S-4 provides an insightful illustration of how rules, regulations, bylaws, instruments, notices, and other forms of subordinate legislation can all be enacted under the same statute by one or more delegates. Section 12 empowers the Alberta Securities Commission to make bylaws governing the management of the Commission itself. The Executive Director and the Commission have powers to issue Orders under the act on various subjects. Section 223 gives the Lieutenant Governor in Council (provincial cabinet) wide powers to enact regulations governing securities transactions generally, and section 224 gives the Commission the power to make rules on the same topics governing in section 223. Some of these Commission rules need prior approval of the responsible Minister, others do not.

There are at least three takeaways from this overview: (1) one should resist the urge to identify subordinate legislation simply by its name and be wary of anyone who gives an overly general or simplistic take on explaining delegated legislative powers – this is definitely a substance over form exercise; (2) the courts themselves struggle with categorizing subordinate legislation, for example in Does the Standard of Review Analysis Apply to a Vires Determination of Subordinate Legislation? I noted the court used the odd term of ‘administrative legislation’ to describe a code of ethics enacted by a professional college under its governing statute, and in Auer the Court of Appeal distinguishes ‘true regulations’ enacted by the Governor in Council (federal Cabinet) from regulations made by other delegates (at para 34) while in Portnov the Federal Court of Appeal  categorizes all regulations as ‘administrative decision-making’ (at para 23); (3) the extensive range of delegated authority means it is crucial to identify whether a decision made under delegated authority has the necessary legislative character to be subordinate legislation.

Unfortunately, the jurisprudence gives us less than adequate clarity on how to identify whether something is legislative or not. Canadian courts have identified factors which suggest an instrument is legislative, including: (1) enacted pursuant to a power granted in statute; (2) contains provisions which set a general norm or standard to be followed; (3) uses language that demonstrates an intention to be mandatory; (4) published or otherwise available to the public; and (5) creates sanctions for non-compliance with its provisions (see e.g. Greater Vancouver Transportation Authority v Canadian Federation of Students — British Columbia Component, 2009 SCC 31 (CanLII) at paras 58 – 66). These factors are less precise in application than they appear. For instance, what constitutes a ‘general norm’ or ‘an intention to be mandatory’ or ‘available to the public’ is often a matter of interpretation.

Conflicting Authority

The world of subordinate legislation is a complicated and dimly lit tapestry, thus it is perhaps not a surprise that judicial review on the vires of subordinate legislation is somewhat of a mess (see John Mark Keyes, Judicial Review of Delegated Legislation: The Long and Winding Road to Vavilov, 2020 CanLIIDocs 3679). There are generally two lines of authority on the standard of review question. One group of cases holds the position that a standard of review analysis does not apply in this circumstance. The leading decision here is Katz Group Canada Inc v Ontario (Health and Long?Term Care), 2013 SCC 64 (CanLII) (Katz), which sets a very deferential approach in a vires determination with a presumption of validity, no inquiry into the policy merits of the enactment, and a requirement that a challenger establish that the enactment is irrelevant, extraneous, or completely unrelated to the purpose in the governing statute (at paras 24 – 28). Katz was followed in Auer and TransAlta Generation. The doctrinal problem with Katz is that the Supreme Court did not situate this decision within the broader context of administrative law and the principles of judicial review. In other words, while Katz did not apply a standard of review analysis, the decision fails to explicitly state why it did not do so.

In Judicial Review on the Vires of Subordinate Legislation, I wrote about West Fraser Mills Ltd v British Columbia (Workers’ Compensation Appeal Tribunal), 2018 SCC 22 (CanLII) (West Fraser Mills), one of the decisions in the line of authority which asserts a standard of review analysis does apply to a vires determination of subordinate legislation. West Fraser Mills was one of my favourite teaching cases in Administrative Law prior to Vavilov because of how the majority, concurring, and dissenting opinions split on how to select and apply the standard of review. The impugned subordinate legislation in West Fraser Mills is a regulation enacted by the BC Workers’ Compensation Board that imposes a duty on owners of a forestry operation to ensure that operations are planned and conducted in accordance with safe work practices. The Supreme Court characterizes the enactment of the regulation as an exercise of delegated administrative power by the Board, and holds that judicial review of that exercise of power is governed by administrative law principles and a standard of review analysis (West Fraser Mills at para 8). Whereas the majority concluded the applicable standard of review is reasonableness, Justices Russell Brown and Suzanne Côté applied a standard of review analysis but notably departed from the majority by distinguishing a vires determination from the judicial review of other forms of administrative decision-making (adjudicative or otherwise) – concluding that the applicable standard of review in this matter was correctness. Read together, Justices Brown and Côté held that the question of whether a statutory delegate is authorized to enact subordinate legislation is manifestly jurisdictional and that a distinction should be made between judicial review on an exercise of legislative and non-legislative power by a statutory delegate. The primary point being that the exercise of legislative power is distinct from other types of delegated or administrative decision-making and should presumptively not attract judicial deference. In a review of delegated legislative authority, Justice Côté held that deference should be afforded where there is a means for democratic accountability, such as public hearings in a bylaw enactment or a legislated notice and comment process (West Fraser Mills at paras 64 – 66).

At this point, the mess becomes obvious. The Katz decision articulates a hyper-deferential approach to judicial review of a regulation but, as Justice Côté hints at in West Fraser Mills, the Supreme Court in Katz arguably engaged in a very intrusive review of regulations enacted by the Ontario Lieutenant Governor in Council. West Fraser Mills does not overrule Katz, and the majority referenced Katz in its application of reasonableness (West Fraser Mills at para 12), however Justice Côté in dissent referenced Katz as an example of correctness review (West Fraser Mills at paras 67 – 70)! This mess was completely avoided by the Supreme Court in its overhaul on standard of review with the 2019 Vavilov decision by not explicitly addressing how the Vavilov standard of review framework would apply to a review on the vires of subordinate legislation.

Not surprisingly, courts post-Vavilov have split on this question. The Federal Court of Appeal’s 2021 Portnov and 2022 Innovative Medicines Canada decisions ruled that the Vavilov framework applies, and the two decisions subject to this comment – Auer and TransAlta Generation – have ruled that Vavilov does not apply. Commentators have likewise taken sides. In Reviewing Regulations Post-Vavilov: Ecology Action Centre v Canada (Part II) my colleague Professor Martin Olszynski (along with Mark Mancini) endorsed the Portnov decision and its conclusion that the Vavilov framework replaced the Katz analysis. Similarly, Professor Paul Daly has argued here that the Vavilov standard of review framework applies. Sara Blake, on the other hand, agrees with the decision in Auer to reject an application of Vavilov and endorses Katz here. Much of these differences are informed by how one categorizes administrative or delegated decision-making and by how these categories fit within the separation of powers between the legislative, executive, and judicial branches of government.

TransAlta Generation and Auer

The applicant in TransAlta Generation challenged the vires of ministerial guidelines issued under sections 322 and 322.1 of the Municipal Government Act, RSA 2000, c M-26. The dispute concerned the assessment of coal-fired power generation facilities, and specifically that the guidelines excluded a consideration of legislation on the phase-out of these facilities by 2030 for the purposes of depreciation. The applicant’s position included arguments that the guidelines were inequitable, discriminatory, and ultra vires the Minister. The Chambers Justice held that a Katz analysis could be applied within the Vavilov framework of a reasonableness review (sort of akin to the majority in West Fraser Mills), and accordingly ruled the guidelines were reasonable and within the authority of the Minister. The Court of Appeal upheld the Chambers Justice’s decision in result (re vires), but ruled that while the Chambers Justice purported to apply Vavilov, her reasoning demonstrated a Katz analysis:

Although the parties take no issue with the chambers judge selecting the reasonableness standard to review the 2017 Linear Guidelines, it is unclear how this standard of review informed her analysis. What the parties – and the chambers judge – effectively label as a review of the 2017 Linear Guidelines for “reasonableness” is more precisely characterized as a matter of statutory interpretation guided by principles that preserve the separation of powers between the legislative, executive, and judicial branches of government by reminding courts not to overstep and assess whether regulations made by delegated lawmakers are “necessary, wise, or effective in practice”: Katz at paras 24-28. (TransAlta Generation at para 40)

The Court of Appeal ruled that a vires review of regulations must follow Katz and that this approach was not replaced or modified by Vavilov (TransAlta Generation at para 46). The Court’s decision in Auer gives a more thorough explanation for this conclusion.

The applicant in Auer challenged the vires of federal child support guidelines issued by the Governor in Council under the Divorce Act, RSC 1985, c 3 (2nd Supp). When the guidelines were brought into force in 1997, they replaced a discretionary judicial determination on child support amounts payable, which was criticized for inconsistencies and unpredictability in divorce proceedings. The applicant’s position included arguments, based on expert evidence, that the federal guidelines lead to a disproportionate obligation not in accordance with the principle of joint financial responsibility under the Act. Like in the initial proceedings of TransAlta Generation, the Chambers Justice in Auer applied a Katz analysis to give a ‘tempered’ reasonableness review under Vavilov (Auer at para 3).

The Court of Appeal in Auer gives 4 doctrinal reasons for why the Vavilov standard of review framework does not replace Katz on a vires review of regulations:

For the reasons below, I am of the view that the test articulated in Katz Group remains the appropriate test to apply when the vires of Governor in Council regulations is being challenged; this test has neither been overtaken nor modified by Vavilov. I say this because:

• • There is a distinction between administrative decision-making and legislative action; specifically, the creation of law through promulgation of regulations by the Governor in Council. Enacting a regulation is therefore not a “decision” in the Vavilov sense and applying the Vavilov reasonableness approach is not analytically sound.
  • Nothing in Vavilov evidences an intention to overrule or modify the Katz Group test; to conclude that Vavilov did so implicitly is inimical to the Supreme Court’s stated goals and reasoning in Vavilov.
  • The Vavilov reasonableness standard of review does not apply neatly to a vires challenge of Governor in Council regulations; the reasoning in Vavilov is fundamentally unsuited to the review of regulations.
  • Vires is not a question of jurisdiction as defined by Vavilov and remains a free-standing ground of review.

Applying the Katz Group test, the Guidelines are legal, and the appeal must be dismissed. (Auer at para 7)

I agree with most of this. The Vavilov framework is a poor fit for a vires determination of subordinate legislation because it fails to properly explain and specifically distinguish legislative decision-making (the reasoning in Vavilov has almost nothing about legislative decision-making) and its overall approach on justification is somewhat naïve to the fact that a full record of decision for subordinate legislation will rarely be available in judicial review. Federal regulations are maybe an exception to this problem if you consider a Regulatory Impact Assessment Statement (RIAS) to be a full record, but this is debatable (see section 7.2 of the federal Policy on Regulatory Development for a discussion of RIAS). Accordingly, in most situations the record for a judicial review on subordinate legislation will need to be supplemented, which will almost certainly lead to process quagmire. Also, while not directly on point here, the selection of a standard of review under Vavilov relies heavily on a consideration having little or no application to delegated legislative decision-making – legislative intent and statutory rights of appeal. It is nonsensical that a legislature would ever provide for a right of appeal to the judiciary on the enactment of subordinate legislation.

However, that is not to say that Katz is perfect, and I think Auer loses much of its persuasion because the decision adheres to formalism and a strict separation of powers that was glossed over in Katz and does not exist in Canada. Auer also fails to properly grapple with the complexities of delegated lawmaking.

The Court of Appeal in Auer begins its defence of Katz by distinguishing a vires determination from the ‘question of jurisdiction’ extinguished in Vavilov (Auer at paras 36 – 41). While acknowledging that the terms ‘vires’ and ‘jurisdiction’ are often used interchangeably by commentators and courts, the Court of Appeal asserts that Vavilov’s references to a jurisdictional question relate exclusively to administrative decisions, not legislative ones; therefore, it follows that a vires determination on regulations is not eliminated or even addressed by the Supreme Court in Vavilov.  I agree. In The Great Divide on Standard of Review in Canadian Administrative Law, I wrote that the Supreme Court’s apparent obsession with eviscerating ‘questions of jurisdiction’ in standard of review analysis was odd because a concern with limits of power and the ‘jurisdiction’ of statutory decision-makers more or less underlies the entire world of judicial review in administrative law (See also more recently David Jones, Vavilov: What it does, and what it does not do at 9 – 10).

The legislative – administrative decision-making dichotomy is at the core of how both Auer and TransAlta Generation distinguish Vavilov from applying to a vires determination of regulations. The Court of Appeal observes that:

1. the narrative in Vavilov is focused almost entirely on a review of an adjudicative or allocative administrative decision;
2. there are few references to the judicial review of legislative power in Vavilov;
3. all references are ancillary and do not speak directly to the Katz

(TransAlta Generation at paras 44, 45; Auer at paras 42 – 46).

The Court of Appeal also remarks that the Vavilov considerations on how to assess reasonableness in a review of statutory power are aligned with administrative decision-making and do not fit well with legislative decisions:

1. the enactment of regulations is not based on the submissions or evidence before a decision-maker;
2. there is no application or hearing process in legislative decision-making;
3. there is no specific or individualized factual context in the enactment of a regulation;
4. reasons for why a regulation is enacted or the reasoning process underlying that enactment are neither provided nor required in law.

(Auer at paras 66 – 68, 72 – 74).

Accordingly, the Court of Appeal concludes that the Vavilov framework is “. . . so fundamentally unsuited to examining the vires of regulations that the Supreme Court of Canada could not have intended to silently overrule the long-standing Katz Group test.” (Auer at para 78). While I agree with this conclusion, I think the Court of Appeal overstates its observations on the character of delegated law-making.

This legislative – administrative dichotomy is also closely aligned with a strict separation of powers between the legislative, executive, and judicial branches of government. As the Court of Appeal in Auer explains:

As explained in Ontario v Criminal Lawyers’ Association of Ontario, 2013 SCC 43 at para 28, the executive, legislative, and judicial branches of government each play distinct and critical roles in our constitutional democracy:

The legislative branch makes policy choices, adopts laws and holds the purse strings of government, as only it can authorize the spending of public funds. The executive implements and administers those policy choices and laws with the assistance of a professional public service. The judiciary maintains the rule of law, by interpreting and applying these laws through the independent and impartial adjudication of references and disputes, and protects the fundamental liberties and freedoms guaranteed under the Charter.

No one branch should overstep its bounds, but rather, each must show proper deference to the legitimate spheres of the others.

The separation of powers has been described as the backbone of our constitutional system: R v Chouhan, 2021 SCC 26 [Chouhan] at para 130, per Rowe J:

While the legislature “chooses the appropriate response to social problems, makes policy decisions and enacts legislation”, the judiciary “interprets and applies the law, … acts as judicial arbiters” and ensures that laws and government action conform to constitutional norms” [citation omitted].

This comment encapsulates the legislative and policy making nature of the enactment of regulations – an act incidental to the legislative process. (Auer at paras 54 – 56)

This formalism led the Court of Appeal to some erroneous distinctions in Auer, such as that ‘administrative decision-making’ is not ‘law-making’ (Auer at paras 52, 79), that there is such a thing as ‘regulations’ but then there are also ‘true regulations’ (Auer at para 34), and that “[e]nacting a regulation is not a ‘decision’ in the Vavilov sense.” (Auer at para 79). Each of these distinctions is either easily flipped or seems untenable (e.g. the distinction between a ‘true’ regulation and other regulations). As well, the strict separation of powers described by the Court is, in practice, a murky separation at best. In particular, the executive branch clearly exercises legislative power, and does more than merely implement policy direction in modern governments (as I would suggest both Auer and TransAlta Generation forcefully illustrate!).  And the judiciary is not merely an adjudicator of disputes and interpreter of legislation; over the years, direction on many significant social policy issues has come from the judicial branch (e.g. Carter v Canada (Attorney General), 2015 SCC 5 (CanLII), and medical assistance in dying).

It seems to me, this formalism led the Court of Appeal to overstate things like: the enactment of regulations is not based on “evidence” (Auer at para 66), regulations have a universal application (ibid), there are no “facts” involved in passing regulations (Auer at para 67), regulations are internally generated by the executive branch (ibid), reasons are never given for regulations (Auer at para 72). While these observations may be accurate in some instances, they definitely cannot be made as absolute claims. As I explained above, the world of delegated law-making and subordinate legislation is far too nuanced to be explained summarily in such terms.

This suspect foundation is the basis for the Court of Appeal’s distinction and conclusion on three forms of decision-making and the correct approach to judicial review. In descending order of deference owed:

1. Regulations enacted by the political executive (Cabinet or a minister) are reviewable under the Katz approach and its presumption of validity. This is ‘true’ delegated legislating, without a factual context.  The separation of powers precludes the judiciary from reviewing the policy merits of these regulations outside of a constitutional question (Auer at para 81; TransAlta Generation at para 53);
2. All other subordinate legislation enacted by statutory delegates, such as administrative agencies, are reviewable under a modified version of Vavilov’s reasonableness. Presumably this review looks somewhat along the lines of the pre-Vavilov cases that applied a standard of review to subordinate legislation, but no clear guidance has been offered and recall how split the Supreme Court was in West Fraser Mills (Auer at para 82; TransAlta Generation at para 49);
3. All non-legislative or administrative decisions made by the political executive or statutory delegates are reviewable under a Vavilov standard of review framework.
   This seems straightforward enough so long as you can persuasively distinguish a legislative decision from an administrative one. Courts often address this question in administrative law because the common law doctrine of procedural fairness does not apply to a legislative decision (Knight v Indian Head School Division No 19, [1990] 1 SCR 653), but it is not always an easy task and is often contested (as it was in TransAlta Generation, see paras 88 – 98).

Conclusion

Without a full understanding of delegated lawmaking, it really is a hopeless endeavour to arrive at an approach to judicial review that respects both: (1) the role of the judicial branch to administer and uphold the principle of legality and the rule of law in supervising the exercise of legislative power by the executive branch; and (2) the power to legislate being inherent to the legislative branch and its delegates (the executive).

Full Vavilov on a vires determination of subordinate legislation is not the answer. The Vavilov framework gives insufficient attention to the exercise of legislative power.  Subordinate legislation cannot properly be situated with adjudicative and other administrative decision-making under a one-size-fits-all umbrella of judicial review in administrative law. Context matters. The process of decision-making and the substance of decisions can be, and often is, very distinct in legislative and adjudicative settings.

No Vavilov on a vires determination of subordinate legislation is not the answer. The Katz approach is wholly inadequate in relation to the principle of legality and the exercise of legislative power. It is well past time for Canadian courts to address the pervasive character of delegated lawmaking in our government, the democratic deficit inherent in this approach to legislating, and the challenges this poses for the principle of legality.

Partial Vavilov on a vires determination of subordinate legislation is likely the answer. The overall direction of a ‘robust’ reasonableness review under Vavilov does have potential, however it remains to be seen how this type of review can implement the principal of legality, respect the authority of the executive branch to legislate, and fit it all into the mechanics (and constraints) of a judicial review process.

This post may be cited as: Shaun Fluker, “Judicial Review on the Vires of Subordinate Legislation: Full Vavilov, Partial Vavilov or No Vavilov?” (February 6, 2023), online: ABlawg, http://ablawg.ca/wp-content/uploads/2023/02/Blog_SF_Auer.pdf

To subscribe to ABlawg by email or RSS feed, please go to http://ablawg.ca

Follow us on Twitter @ABlawg