Post-Truth, Redux

A faithful application of Vavilov reasonableness review exposes the rot at the core of Canada’s administrative law

Co-blogger Mark Mancini has already posted on the Federal Court of Appeal’s recent decision in Alexion Pharmaceuticals Inc v Canada (Attorney General), 2021 FCA 157. He argues that it is a good illustration of how courts should review administrative decisions on the reasonableness standard, following the Supreme Court’s instructions in Canada (Minister of Citizenship and Immigration) v Vavilov, 2019 SCC 65. I agree with Mark’s analysis, so far as it goes: as a rigorous application of Vavilov that rightly emphasizes legal constraints on administrative decision-making, Justice Stratas’ reasons for the Court in Alexion are excellent. (In fact, let me highlight an additional passage that Mark does not mention: Justice Stratas notes, rightly, that administrators must interpret statutes “in a genuine, non-tendentious, non-expedient way … Result-oriented analysis is no part of the exercise”. [37] Amen!)

But, in my view they are also an excellent illustration of the considerable flaws of the Vavilov framework, with its insistence on the centrality of administrative reasons on all issues subject to the reasonableness standard of review, including issues of statutory interpretation. Indeed, Alexion illustrates the fundamental soundness of the approach taken in the case that is the great bogeyman of Canadian administrative law: the House of Lords’ Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147. The concurring judges in Vavilov accused the majority of following Anisminic. If only!


As Mark explains in more detail, Alexion reviewed a decision by the Patented Medicine Prices Review Board that the company was selling a product “at a price that, in the Board’s opinion, is excessive” (s 83 of the Patent Act). The Court of Appeal invalidated the Board’s decision, holding that it did not explain its reasoning on key issues, including the interpretation of s 85 of the Patent Act, which sets out the criteria the Board must apply in deciding whether the price of a patented medicine is “excessive”. As Justice Stratas notes,

[a]t best, on this point the Board obfuscated, making it impossible for a reviewing court to know whether the Board has helped itself to a power it does not lawfully have. By obfuscating, the Board has effectively put itself beyond review on this point, asking the Court to sign a blank cheque in its favour. … 

[T]he Board may have helped itself to powers the statute has not given it. The absence of a reasoned explanation on certain points means that we cannot be more definitive than that. [44]-[45]

Justice Stratas notes that the Board appears to have found the pricing of Alexion’s product unreasonable, and expresses his “fundamental concern … that the Board has misunderstood the mandate Parliament has given to it under s 85. At a minimum, a reasoned explanation on this is missing“. [48; emphasis mine] And further:

Section 85 speaks of “excessive” pricing, not  “reasonable” pricing. The two seem much different. If in fact they are not different, in the circumstances of this case the Board had to explain why. Nowhere does the Board do so. [52; emphasis mine]

If I may paraphrase Justice Stratas, he is saying: it looks like the Board is doing something it’s not supposed to be doing under the statute, but hey! maybe it’s not do these things, or maybe it can do these things after all ― and we, the Federal Court of Appeal, can’t know for sure. The suggestion here ― that, absent good quality reasons given by the administrator, a reviewing court cannot say whether the administrator, in Justice Stratas’ eloquent words, “helped itself to a power it does not lawfully have” ― is entirely consistent with Vavilov. There the majority insisted that

the focus of reasonableness review must be on the decision actually made by the decision maker …  The role of courts in these circumstances is to review, and they are, at least as a general rule, to refrain from deciding the issue themselves. Accordingly, a court applying the reasonableness standard does not ask what decision it would have made in place of that of the administrative decision maker … conduct a de novo analysis or seek to determine the “correct” solution to the problem. [83]

On this approach, Justice Stratas and his colleagues are not supposed to come to their own view of the meaning of s 85 and verify the Board’s compliance with it. They are confined to assessing the Board’s explanations as to whether it has complied. Absent an explanation, the exercise fails. Vavilov is an improvement over the earlier cases in that, when such failures occur, it allows the reviewing court to stop there and send the matter back to the administrator for a do-over, instead of making up an explanation and deferring to it. (See Mark’s post for more on this).


But to say that Vavilov improves over what I once described as a post-truth jurisprudence requiring judges to play chess with themselves and contrive to lose is not to say much. In fact, Vavilov does not even leave post-truth jurisprudence behind. For how else should we think of a requirement that judges ― of an appellate court, no less ― insist that they “cannot be definitive” about the interpretation of a statutory provision and about whether an administrator “helped itself to a power it does not lawfully have” ― which is to say, exceeded its jurisdiction (there, I said it) in applying that provision?

The truth is that judges can be definitive on such things. The truth is that Justice Stratas has much to say about the meaning of s 85 and the way in which it has to be applied, as well as the more general principles of statutory interpretation (see, in particular, his important caution that “[t]he authentic meaning of the legislation … is the law, not what some politicians may have said about it at some place, at some time, for whatever reason”). [53] (I recently addressed this point here.) The truth is that, as Justice Stratas notes, “[o]ver and over again, authorities have stressed that the excessive pricing provisions in the Patent Act are directed at controlling patent abuse, not reasonable pricing, price-regulation or consumer protection at large”. [50] A jurisprudence that requires a court to assert that, notwithstanding all of this, an administrative tribunal might somehow explain all that away, and show that when it said “reasonable” it meant “excessive”, and that when it “disregarded most of the … authorities”, [51] it still complied with the law, is the jurisprudence of la-la-land.

In reality, the Board’s decision has all the appearances of a textbook example of what Lord Reid in Anisminic described as an administrative tribunal having “misconstrued the provisions giving it power to act so that it failed to deal with the question remitted to it and decided some question which was not remitted to it”. When a tribunal does so, even though in a narrow sense “the tribunal had jurisdiction to enter on the inquiry”, it loses jurisdiction in a broad sense, and the resulting decision is a nullity. Canadian courts should be able to say so ― which means that they should be free, contra Vavilov, to “decide the issue themselves”, without waiting, or even affecting to wait, to be instructed by administrators who lack the legitimacy, the independence, and the competence to speak on questions of law with any real authority.

Why is it that we can’t have nice things? An important part of the problem is the fusion, in Canadian administrative law, of what in the United Kingdom (and New Zealand) are known as legality review and reasonableness review into a (supposedly) unified category of merits review. To make things worse, the Supreme Court remains committed to an oversimplified approach to merits review, such that it almost always has to be conducted on the same reasonableness standard. The reasons-first approach may be suitable for review of fact- or policy-based administrative decisions, but applied to issues of statutory interpretation it leads to Alexion-style absurdity.

What makes Alexion even more galling, though, is the nature of the administrative body it concerns. And that’s not only, and perhaps even so much, that, pursuant to s 91 of the Patent Act the Board’s members can legally be the first five strangers the Minister of Health meets on the street one day ― or hacks. (As I wrote this, I thought I’d look up the Board’s actual membership, in the hope of being able to add a disclaimer to the effect they are all, in fact, wise and experienced experts. Only, there doesn’t seem to be any information about them on the Board’s website. Of course that doesn’t prove that they actually are hacks, let alone people the Minister met on the street, but one might have thought some transparency was in order. UPDATE: Mea culpa. The information is there, however counter-intuitive its presentation may seem to me. The members’ bios are here.)

Worse is the fact that the Board acts as both prosecutor and judge in the cases it handles, the separation of powers be damned. This par for the course in the administrative state, to be sure ― but no less pernicious for all that. I note, for the sake of completeness, that it is “Board Staff” that “filed a Statement of Allegations” against Alexion, rather than Board members ― but staff (pursuant to s 93(2)(b) of the Patent Act) are managed by the Board’s Chairperson, i.e. one of its members. The Board’s internal “separation of powers” is more sham than ersatz.

Why exactly should the views of this judge-and-prosecutor, this two-headed abomination against due process of law, about the meaning of the statute it is charged with applying be entitled to any regard by actual judges? In Vavilov, the Supreme Court insists that this is to respect Parliament’s intent. But, as I have been saying since my first comment on Vavilov here, the Court ignores Parliament’s direction, in s 18.1(4)(c) of the Federal Courts Act that the federal courts grant relief when administrative decision-makers err in law, which clearly requires these courts to come to their own view about what statutes mean and whether the administrator in a give case has complied with the law. In this way too, Vavilov perpetuates Canadian administrative law’s disregard for truth.


In case this needs to be clarified, none of the foregoing is a critique of Justice Stratas and the decision in Alexion. As I said above, I think that the decision is about as good as it could have been while being a faithful application of the Vavilov framework. If the Board takes what Justice Stratas seriously, it will make a much better, and most importantly, a lawful decision next time. It is the framework itself that is rotten.

But the rot set in four decades ago, and no judge of the Federal Court of Appeal can solve them ― even one who has made Herculean efforts to, like Justice Stratas. Perhaps even the Supreme Court cannot fully undo the damage it has inflicted on our law when it turned away from the Anisminic path and waded into the dark forest of deference to the administrative state. But if Alexion illustrates the possibilities ― and the limits ― of what the Supreme Court accomplished in Vavilov, and I think it does, then one has to conclude that the Supreme Court hasn’t tried very hard at all.

Alexion: No Blank Cheques Here

In Alexion Pharmaceuticals Inc v Canada (Attorney General), 2021 FCA 157, the Federal Court of Appeal clarified the law of judicial review post-Vavilov (particularly as it applies to reasonableness review) and set out an important reminder: administrators are not a law unto themselves. In order to make sure that this is the case, particularly in situations of legislative interpretation, administrators must explain their decisions. They must do so in a way that engages with the statute under interpretation. In this way, Alexion says something important: when administrators interpret statutes, there is only so much of a margin of appreciation. They must deal with the law.

I first describe the controversy in Alexion and the Court’s holding. Then I outline why this decision is a landmark one for the post-Vavilov world.

**

Alexion is a pharmaceutical company that produces a drug called Soliris. The Patented Medicine Prices Review Board had to decide whether Alexion priced Soliris excessively under the Patent Act. The relevant section is s.85 (1), which lists a number of factors that the Board must consider to make a determination. One of the factors is “the prices at which the medicine and other medicines in the same therapeutic class have been sold in countries other than Canada” (s.85(1)(c)). Only after a consideration of these factors can the Board turn its attention (if necessary) to s.85(2), which asks the Board to also consider “the costs of making and marketing the medicine and any other factors it considers relevant.”

As the Court aptly notes, “Section 85 is the law. The Board’s analysis should start with the law. Whatever the Board does must be consistent with the law” (Alexion, at para 34). The Board, in making its excessive pricing decision, noted that it is charged with determining “the relevance and weight of each factor” in s.85 (Alexion, at para 43). The Board concluded that Soliris was priced excessively, largely because it was priced more than the lowest international price in a list of seven comparator countries (Alexion, at para 3).  Put differently, out of all the seven countries, the Board found Soliris to be priced excessively because it was not the cheapest option. This was despite the fact that the Board’s own guidelines suggested that, normally, “the highest international price” is a key comparator (Alexion, at para 57). In reaching this conclusion, the Board seemingly followed a standard of “reasonableness”: because Soliris is price higher than one of the comparator countries, the Board implicitly concluded that the price of Soliris is unreasonable (see Alexion, at para 51).

For the Court, Stratas JA concluded that the Board failed to properly justify its decision with reference to the statute at hand: Alexion, at para 64, 66. The Court made a number of important comments justifying this decision:

  1. Prior to Vavilov, “…the Supreme Court instructed us to do our best to try to sustain the outcomes reached by administrators” which included “reviewing courts [picking] up an administrator’s pen and [writing] supplemental reasons supporting the administrators’ outcomes” (Alexion, at para 8). This “ghostwriting” was, as is evident, “antithetical to the reviewing courts’ role as an independent reviewer” (Alexion, at para 8).
  2. In this sense, there is a clear relationship between reasons and outcome on judicial review (Alexion, at para 28 et seq). While Vavilov speaks of outcome and reasons as separate, there are many cases where the reasoning on a particular legal question will lead to an illegal outcome; for example, in this case, “certain words the Board used suggest that it went beyond its permissible statutory mandate by regulating the reasonableness of pricing, rather than preventing abusive pricing…” (Alexion, at para 11).  In this case, when the Board spoke of “reasonableness” rather than abusive pricing, “[i]t may be that the Board was trying to reach an outcome that on the facts and the law was not reasonably open to it” (Alexion, at para 32).
  3. The failure of explanation in this case arose on a few different fronts:
  • The Board utterly failed to deal with the most important and central restriction on its authority: s.85 of the Patent Act. We  know that in paras 120-122 of Vavilov the Court notes that “the merits of an administrative decision maker’s interpretation of a statutory provision must be consistent with the text, context and purpose of the provision,” and that the decision-maker must demonstrate that its alive to these “essential elements.” This is because “the governing statutory scheme is likely to be the most salient aspect of the legal context relevant to a particular decision” (Vavilov, at para 108). So when the Board adopted a standard of reasonableness rather than addressing the actual point of the statute—set out in s.85 and the associated case law—it transgressed its authority.
  • The Board’s failure to explain its departure from its own Guidelines was problematic from a reasonableness perspective. While Guidelines adopted by the Board cannot supercede an analysis based on s.85 itself, they can validly guide discretion. Here, the Board did not explain why it did not follow its own Guidelines, which stressed the highest price comparator country.

**

There is a lot packed into Alexion, but I think it is worth noting the various things the Court does with Vavilov, especially when it comes to the reasonableness standard.

First, the Court arguably doubles down on the statute as the most important restraint on administrative power. Many of us who read Vavilov for the first time in December 2019 fastened onto paras 108-110 (and also paras 120-122) of that decision as quite important. Those paragraphs hardened a cardinal rule of administrative interpretations of law: it is the statute that the administrator is interpreting (its text, context, and purpose) that cabin the discretion of an interpreting administrator. Now, how this happens is where the rubber meets the road. But the fact that the statement was made by the Supreme Court—and that it is adopted wholeheartedly by the Court of Appeal in this case—is promising.

There are, of course, different ways that a court can ask an administrator to abide by the terms of its statute, and these ways can be more or less interventionist. Alexion is somewhat reminiscent of another case decided post-Vavilov, Richardson. I blogged about that case here. While the comments made by Nadon JA in that case were obiter, they staked out an even more radical understanding of Vavilov’s paragraphs 108-110 and 120-122. In that case, the administrator at hand erroneously applied the “implied exclusion” rule of interpretation, which the Supreme Court has held is insufficient as the sole basis on which to understand the meaning of statutory provisions (see Green, at para 37). Imposing the Supreme Court’s method of interpretation, particularly with regards to particular canons, is one way to force an administrator to abide by a statute. Another, more general and less stringent way, is what Stratas JA did in Alexion. There the Board misapprehended its own statutory purpose and failed to actually deal with the overriding goal of s.85: excessive & abusive pricing. It also ignored many of the factors set out in s.85(1). This is just a different way of getting at paras 108-110 of Vavilov: the Board failed to address its statute under the governing approach to statutory interpretation.

The fact that the Court in Richardson and Alexion did the same thing in different ways is perhaps indicative of a challenge with Vavilov. The decision says a lot, not all of it always internally consistent. Specifically, the challenge going forward with this rather legalistic vision of reasonableness review is how it meshes with the deference that is built-in to the Vavilov framework. Vavilov makes clear at various points that administrators are not asked to engage in a formalistic interpretation exercise (para 119), and that ‘[a]dministrative justice’ will not always look like ‘judicial justice’…” (para 92). Accordingly, as Professor Daly notes, “some portions of Vavilov are liable to become battlegrounds between different factions of judges, those who favour more intrusive review on questions of law in one camp, their more deferential colleagues in the other” (at 15). One could conceive, as Professor Daly does, of Richardson as “betraying a favouritism for an interventionist standard of reasonableness review on issues of statutory interpretation” (at 14).

However, I would say that Alexion and Richardson are of the same ilk, different points on a similar spectrum. Both are directed towards subjecting administrators to legal requirements, but Alexion does so in a more general way, faulting the administrator for failing to address the relevant statutory purpose (among other things). Richardson does the same thing in a more specific way, faulting an administrator for applying a proper tool of interpretation to the exclusion of the statutory purpose. Both, in my view, are plausible views of Vavilov.

Methodologically, there are other important elements of Alexion. One element is the connection that Stratas JA draws between reasons and outcome. Vavilov speaks of reasons and outcome as separate things, but in reality, they are probably intrinsically connected in at least some cases. Alexion provides a good example. In many cases, it was simply impossible for the Court to determine whether the Board had ventured an opinion on the core legal issue at play in the case. Where the Board did offer an opinion, it cast its decision in terms of the wrong legal standards.  This led it astray, and it was led astray because its reasoning failed to glom onto the important part of the entire thing: the statute.

This leads to a final point about Alexion. Thank goodness we no longer need to worry about courts coopering up deficient decisions under the Nfld Nurses line of cases. As the Court in Alexion mentions, this decision could have gone a very different way under pre-Vavilov case law. The Court would have asked itself to supplement reasons for decision instead of supplanting them.  But as the Court notes, “[m]any of us recoiled at this” (Alexion, at para 9). Why? Because it offends the principle of legality, fundamental to the administrative law system, for a court to uphold a decision that is legally flawed. Of course, deference sometimes asks us to abridge the principle of legality in a strict sense; but there are extremes, and a court making a decision for an administrator is to my mind (and, apparently the mind of the Supreme Court) a bridge too far. As the Court in Alexion says, there are no blank cheques in the law of judicial review (Alexion, at para 44).

All told, Alexion is an important recap of developments post-Vavilov. Particularly on the application of the reasonableness standard, the Court moves the needle in important ways.

“Administrative Sabotage” and the Ontario Human Rights Tribunal

Recently, Professor David Noll (Rutgers Law) posted a fascinating article called “Administrative Sabotage” on SSRN, forthcoming in the Michigan Law Review. You can view the article here, and Professor Noll wrote a fascinating thread outlining its main arguments. The abstract:

Government can sabotage itself. From the president’s choice of agency heads to agency budgets, regulations, and litigating positions, presidents and their appointees have undermined the very programs they administer. But why would an agency try to put itself out of business? And how can agencies that are subject to an array of political and legal checks succeed in sabotaging statutory programs?

This Article offers an account of the “what, why, and how” of administrative sabotage that answers those questions. It contends that sabotage reflects a distinct mode of agency action that is more permanent, more destructive, and more democratically illegitimate than other more-studied forms of maladministration. In contrast to an agency that shirks its statutory duties or drifts away from Congress’s policy goals, one engaged in sabotage aims deliberately to kill or nullify a program it administers. Agencies sabotage because presidents ask them to. Facing pressure to dismantle statutory programs in an environment where securing legislation from Congress is difficult and politically costly, presidents pursue retrenchment through the administrative state.

[…]

Professor Noll’s paper is a significant contribution, relevant outside of the United States. In fact, as I have written about previously (see Mark Mancini, “The Political Problem with the Administrative State” (2020) 2 Journal of Commonwealth Law 55) the Ford government’s treatment of the Ontario Human Rights Tribunal (OHRT) is a classic example of a government legitimately delaying appointments to stymie the practice of the administrative justice system. Professor Noll has now provided the theoretical and linguistic tools for us to understand this phenomenon in administrative government, even for us in Canada.

***

Noll’s focus is “the sabotage of statutory programs by agencies that administer them” [7]. In this, Noll’s project fits in a rich tradition of public administration scholarship that has studied the various ways in which bureaucrats can undermine policy objectives set by their enabling statutes, through mal -administration, “shirking” or drifting, or sabotage. Sabotage can be defined as bureaucratic action that “deliberately undermines policy objectives of the superiors” (see John Brehm & Scott Gates, Working, Shirking and Sabotage: Bureaucratic Response to a Democratic Republic, at 21).  Sabotage “involves a specific stance on the part of the agency toward the program it administers” and the stance “seeks to eliminate a program [the agency] administers” [8]. Sabotage is thus different from other bureaucratic phenomena, in that it involves a deliberate and intentional sacking from within of the agency’s ability to fulfill its delegated mandate [7]. A classic basic example of sabotage that Noll points out—and that I address in my paper on the OHRT—is “non-appointments”—failing to appoint agency heads, or other important positions, as the case may have it [30].

Administrative sabotage, in either Canada or the United States, is a destructive practice that undermines the legislative choice to delegate to agencies. As Noll says: “Rather than use delegated authority to enforce and elaborate statutory policy, an agency uses that authority to undermine the program it administers. In structural terms, this use of delegated authority is at odds with the principle of legislative supremacy” [10]. Once a legislature has delegated power to an agency, it is a condition of the delegation that the power be exercised according to the enabling statute. Agencies and politicians that fail to live up to these delegated terms—and worse, agencies and politicians that actively undermine them—act inconsistently with the power they have been given. Moreover, they act undemocratically—they undermine the legislative plan & bargain containing the conditions governing the administrative action.

Complicating this conventional picture is the emergence of theories of executive control over the administrative state and the desirability of political control as a constitutional matter. The unitary executive theory in the US, for example, generally holds that all executive power is placed in a President, and it therefore follows that the “executive”—including executive administrative agencies—must be controlled by the President (see, for the nuances, Seila Law). In Canada, we have a parliamentary system, but the gist is similar in at least some respects. Legislatures provide powers to executives and administrative decision-makers to make decisions. Legislatures also structure the relationship between the executive and the administrative state, creating and controlling powers of appointment, for example (see the classic example in Saskatchewan Federation of Labour, 2013 SKCA 61). A strong executive power advocate may claim that that the executive can lawfully engage in sabotage by appointing people who wish to undermine the agency itself. It can do so because the executive is the representative of the people, and thus is the politically legitimate actor, in contradistinction to unaccountable administrators.

In the US, the Trump administration furnished many examples of administrative sabotage, and it mooted the defense of the practice. A prominent example included Mick Mulvaney and the Consumer Financial Protection Bureau (the CFPB). The CFPB is in charge of imposing a variety of consumer financial laws. Mick Mulvaney, appointed the head of the CFPB, had previously indicated that he supported abolishing the CFPB [3]. Of course, by itself this is neither here nor there. But once Mulvaney became the head of the agency, on the conventional picture, he had no discretion to undermine the legislative bargain simply because he disagreed with it in principled. Yet he did so: he “declined to request money to fund the Bureau’s operations; installed “Policy Associate Directors” to shadow bureau chiefs protected by the civil service laws; rescinded, stayed, or delayed major rules on payday lending, overdraft fees, and student loan servicing…” [3]. Mulvaney justified these practices by appealing to the adage of “elections have consequences” [11].

Noll’s paper also explores the various reforms that might be adopted to stop sabotage. Noll shows how courts and Congress have been largely unable to control sabotage. Presidents and courts that have a reflexively anti-administrativist agenda may, in fact, be incentivized to exacerbate and permit administrative sabotage. But as a practical matter, there is another issue: many instances of “administrative sabotage” are simply not amenable to judicial review: “it is simple to invent technocratic explanations for agency actions designed to undermine a statutory program…”, and as such, there are evidential hurdles [13-14]. Noll suggests that specific statutory reforms that might shed light on the question, the goal of these reforms being that the statutory schemes are designed to prevent sabotage—“policymakers should not assume that programs will be administered in good-faith” [50]. Noll suggests statutory appointment qualifications consistent with the Constitution; and, notably for our purposes, endorses the proposition that broad statutory delegations (the norm since the New Deal) encourage sabotage [54].

***

There is much in Noll’s piece to recommend it to Canadians, but I want to focus on just two points: (1) Noll’s conclusions about delegated power; and (2) the case of the OHRT, arguably an example of Noll’s sabotage.

As noted above, and since the New Deal, scholars have argued—and sometimes assumed—that broad delegations of statutory power are desirable. So the old case goes, legislatures simply do not have the time and expertise to consider all the factors when legislating; and particularly in complex fields of regulation, it makes sense to delegate power to so-called expert agencies. As a descriptive matter, this is likely true, and for that reason, it makes sense for legislatures to “trade-off” political control for expertise (as Epstein & O’Halloran once put it).  But this does not speak to the degree to which this should happen. As I wrote in my article on the matter [94], and as Noll essentially argues:

The real problem with executive discretion, then, is not that it abridges independence; but that it has a potential of being misused to undermine the limitations on statutory power that arise in the context of a delegating statute. The goal should be to cabin executive discretion tightly so that it, necessarily, cannot undermine delegated legislative power. Broad delegations, on this understanding, should be avoided.

The point is that the solution to sabotage starts not with depending on the good-faith of administrators (as a previous generation of pro-administrativist scholars did), or depending on the political control exercised by an executive actor (who may have incentives to permit sabotage). Instead, it starts with the legislature slightly increasing the cost of legislating by keeping the possibility of sabotage in mind when legislating, and using its powers to put meaningful limits on delegated powers.

This raises an important point about independence. To simplify, in Canada, the independence of administrative decision-making is parasitic on the degree to which a statute permits that independence (famously, see Ocean Port). Statutes can either liberate or constrict executive control over the administrative state. One way for executives to control so-called “independent” tribunals is for the legislature to vest an appointment power in the executive. Assuming this power is exercised according to the terms of the statute, there is no constitutional objection; while independence of administrative decision-making may be a good in some cases, it is not self-evidently legally required (though see the reading of the caselaw suggested by Ron Ellis in his text, Unjust By Design). And political control by elected actors is desirable in a system of responsible government.

But again, this is only true to an extent. Sabotage is quite different from an executive exercising lawfully delegated powers of control; it is a situation where an executive or agency head may intentionally choose to exercise power it does not have to undermine the power it has been granted. The sin of omission here is not that the executive is simply choosing not to exercise delegated power; it is that the executive is actively using its position to undermine the entire statutory bargain setting up the agency.

And this is exactly what happened in the case of the Ontario Human Rights Tribunal. As late as January 2021, observers argued that “…Ontario’s human rights enforcement system has become dysfunctional” in part because “The final resolution of a claim can now take years for individuals who have experienced discrimination.” The cause of this delay: few of the human rights adjudicators whose tenure is at-pleasure have been replaced. I previously studied this phenomenon as an example of a situation where an executive was failing to implement delegated statutory power. Indeed, the relevant legislation delegates power to the Cabinet, who “shall” make appointments to the tribunal [my paper, at 82]. By failing to do so, the government created grist for the mill of its critics, who asserted—not unreasonably—that the government was intentionally starving the tribunal and delaying the resolution of claims.

The failure here is traceable, ultimately, to the legislature—though the executive undermining of delegated power is the evil to which the legislature should have turned its mind. The legislature enacted the tribunal, and it can rescind its powers tomorrow. But executives do not have that authority, which is why sabotage is undesirable. So, in the OHRT case, by failing to impose timelimits for appointments and a minimum number of members, the legislation grants easily-abused delegated appointment power to the executive. So, as Noll suggests, it was the breadth of delegated power that created the conditions for sabotage.

Canadians should pay close attention to Noll’s article. While there are obvious differences between the Canadian and American administrative states, the phenomenon of sabotage is likely a common evil.

Keeping Out or Stepping In?

When should the courts intervene in internal disputes of voluntary associations?

This is my first post since February. Apologies. Things haven’t been great, and might not improve for some time, but I do hope that, at least starting in July, I will be posting more regularly.

In Ethiopian Orthodox Tewahedo Church of Canada St. Mary Cathedral v Aga, 2021 SCC 22, the Supreme Court considered the scope of the courts’ power to interfere with the decision-making of an unincorporated private association. The dispute involved a congregation and some of its members, whom its authorities excommunicated, apparently without having given them much of a hearing.

Justice Rowe provides a neat summary of his judgment for the unanimous Court:

[C]ourts can only intervene in the affairs of a voluntary association to vindicate a legal right, such as a right in property or contract. Membership in a voluntary association is not automatically contractual. Even a written constitution does not suffice. Membership is contractual only where the conditions for contract formation are met, including an objective intention to create legal relations. Such an intention is more likely to exist where property or employment are at stake. It is less likely to exist in religious contexts, where individuals may intend for their mutual obligations to be spiritually but not legally binding. A voluntary association will be constituted by a web of contracts among the members only where the conditions for contract formation are met. [49]

The idea that courts will only intervene in the face of an alleged violation of a legal right follows from the Supreme Court’s earlier decision in Highwood Congregation of Jehovah’s Witnesses (Judicial Committee) v Wall, 2018 SCC 26, [2018] 1 SCR 750. The rules of contract formation, including the requirement that parties intend to create legal relations, and not merely socially or spiritually binding ones, are longstanding. Here, they lead Justice Rowe to find that

there is nothing that can be characterized as an objective intention to make an offer on the part of [the congregation or its leadership], and nothing that can be characterized as an objective intention to accept on the part of any of the [excommunicated members], or vice versa. [52]

But a strict application of these rules, combined with (or perhaps resulting in) the position that even a self-proclaimed constitution of the association is not necessarily a binding contract, means that the courts will keep out of the internal disputes of voluntary associations ― especially, but not only, religious ones.


People whose opinion means more than mine have been critical of this. Over at Administrative Law Matters, Paul Daly writes

the sledghammer employed in Aga obliterates any judicial enforcement of any terms in a voluntary association’s “Constitution”, even terms relating to the basic procedures for removing individuals from the organization. When read with WallAga represents a significant judicial retreat. So much for the ‘supervisory’ jurisdiction.

Again, Professor Daly’s opinion on such issues is more important than mine. But here’s a note of doubt.

I’m not sure why we should be sorry about the judicial retreat, if that’s what Aga and Wall are. The Supreme Court is clear that when employment or property (or access to resources necessary to earn a living, as in some earlier cases involving expulsions from communes of coreligionists) are at stake, the courts still should intervene. What they are retreating from are disputes about membership or leadership of voluntary associations. But should they be involved in such disputes?

I would venture ― tentatively ― that it is sensible enough to think that they should not. As Justice Rowe explains

The law concerning the formation of contractual relations embodies practical wisdom. Many informal agreements that people undertake do not result in a contract. There are, for example, mutual undertakings between friends … or between members of a household … In neither of these examples do the parties (reasonably understood) intend to be subject to adjudication as to the performance of their commitments or to the imposition of remedies such as damages or specific performance. [21]-[22] (paragraph break omitted)

These rules aren’t only about practical wisdom though. They’re also about autonomy. As Justice Rowe himself says, they allow people who don’t want the law to step into their relationships to keep it at a distance. Arguably, unincorporated voluntary associations often are places where people exercise this kind of autonomy from the legal system. Not always, to be sure. But at least as a matter of presumption and default position, there is something to be said for Justice Rowe’s (and the Supreme Court’s) approach.

At the very least, this is a matter on which reasonable people might disagree, and on which compromise solutions may be available. Professor Daly notes that provincial legislation in British Columbia and Québec enables courts to intervene in the affairs of voluntary associations, including, in the case of Québec’s Code of Civil Procedure, of “groups not endowed with juridical personality”. Perhaps this is a defensible choice, albeit one less respectful of people’s autonomy. But it’s not necessarily the only defensible choice.

In fact, there is a difference between the Québec and British Columbia statutes to which Professor Daly refers. The latter only applies to “societies” which (as it makes clear) must be deliberately incorporated by their members. If people come together and form a voluntary association without incorporating, the default regime articulated in Aga will still apply. The position in New Zealand is similar: the Judicial Review Procedure Act 2016 defines “statutory power“, which is normally subject to judicial review, as, in relevant part, “a power or right … conferred by or under … the constitution or other instrument of incorporation, rules, or bylaws of any body corporate”. A constitution or bylaws of an unincorporated association do not count.

This may be a sensible distinction to draw: a group that goes to the trouble of formalizing its operations by incorporating, and obtains the benefits of incorporation, they submit to closer scrutiny by the courts. Otherwise, they will mostly be left to their own devices, except where legal relationships such as employment or ownership are involved, or where the association takes on some regulatory or quasi-regulatory role. In New Zealand, Electoral Commission v Cameron [1997] NZCA 301; [1997] 2 NZLR 421, involving the Advertising Standards Complaints Board which, although not incorporated, effectively regulated what advertisements could and could not be distributed by the media is an example.

In fact, Professor Daly’s position may not be all that dissimilar. He asks: “What is the point of setting out procedures, rights and obligations in a document made available to members when those procedures, rights and obligations can be discarded at will?” I’m not sure how strong this objection is in Aga (where the members of the congregation, even seemingly high-ranking ones, seem not to have been aware of the relevant documents for years). But Professor Daly’s point, if I understand correctly, is that choices about the degree of formality with which an association organizes itself are important. That makes intuitive sense. But I’m not sure that the line should be drawn at the creation of a “constitution” rather than at incorporation. The latter may be a more formal, and a more easily identifiable, step, and so perhaps a better marker for the courts to refer to.

Note, by the way, that my argument here is not about religious associations in particular. There are, indeed, good reasons for the secular courts to be especially wary of intervening in their disputes, because they are likely to implicate theological considerations. But this is only a specific application of the broader principle of autonomy that is implicated when the law chooses one approach or another to judicial intervention in the private sphere. Other kinds of associations may also have reasons for wanting to keep their workings informal and outside the state’s reach.


And, to repeat, perhaps they shouldn’t be allowed to do that. Perhaps the more interventionist position chosen by the Québec legislature is the wiser one. As I have already said, my views here are tentative. But I think that a debate about first principles ― about whether it is in fact right for the state to assume the role of a supervisor over the internal doings on voluntary associations ― is worth having. By tilting the default position towards non-interventionism and so putting the onus on legislatures to introduce a different set of rules, the Supreme Court’s decision in Aga may spur something like this debate.

For What It’s Worth

University of Toronto professor Richard Stacey recently released an article in the University of Toronto Law Journal (paywalled, which is truly unfortunate), arguing that (among other things) the Supreme Court of Canada’s decision in  Vavilov “affirm[s]” the Supreme Court’s controversial decision in Doré  (340; see also 351). To be specific, Stacey says (340-341):

Read together, and building on a rich body of Canadian case law that came before,  Vavilov and Doré  definitively mark a culture shift way from an outcomes-oriented conception of judicial review toward one that engenders a culture of justification…

…in a culture of justification built on a robust conception of reasonableness, constitutional law and administrative law come together in a unified system of public law.

Stacey also makes a number of other arguments, including: (1) the culture of justification apparently endorsed in the same way in  Vavilov and Doré necessarily and logically excludes correctness review (see pg 349) and (2) so-called “Charter values” act as “justificatory resources” that together bind a unified system of public law, bringing  Vavilov and Doré  together (357 et seq). That is the core of Stacey’s argument: the so-called “unity of public law” thesis draws together a requirement that administrative decisions be justified.

Stacey’s argument is interesting, but ultimately unpersuasive. For one, it treads well-worn territory of “administrative constitutionalism” and “the unity of public law,” theories that—as I will argue—are actually undermined by  Vavilov’s formalism. While  Vavilov does put a focus on justification, Stacey’s article does not deal with the parts of  Vavilov that clearly work against his thesis: the dispatch of expertise as a factor governing the standard of review; and the retention of correctness review based on Rule of Law considerations, among other things. Secondly, Stacey’s article does not engage with key scholarship on this issue post- Vavilov that could both strengthen and undermine his case. Finally, Stacey attaches too much determinacy to Charter values, ethereal things that even their most ardent supporters must agree are relatively indeterminate: perhaps not of this world.

***

Stacey sets out his argument on the first issue (the marriage between Doré  and  Vavilov) boldly: “The foundation of both cases is the same, and both judgments ultimately embrace the same conception of reasonableness” [351]. Stacey cites the Alsaloussi case out of the Federal Court, where the Court relied on  Vavilov in a Doré -type case to give guidance “on what a reasonable decision looks like” [351]. To Stacey, this case—and the theory—reinforces “how the two judgments help to draw administrative and constitutional law together in a single, unified system of law” [352].

As I have outlined previously, I do not see how this is the case, even on the terms of the reasonableness standard. Doré  was positively unclear about what administrators should do when faced with a Charter claim, beyond saying that an administrative decision-maker “balances the Charter values with the statutory objectives” (Doré , at para 55 et seq). The only question for a judicial review court is whether “in assessing the impact of the relevant Charter protection and given the nature of the decision and the statutory and factual contexts, the decision reflects a proportionate balancing of the Charter protections at play” (Doré , at para 57). But, unlike the Oakes test, and unlike  Vavilov’s list of constraints, the Doré -line of cases do not provide any guidance on how courts should conduct the proportionality analysis.  Vavilov provides a far more robust and detailed schema of reasonableness than Doré  does, and so to equate these cases on this front is ultimately unpersuasive. The similarity on the reasonableness front—if it exists—is cosmetic at best.

Stacey also does not address why  Vavilov’s comments on constitutional issues do not demand a correctness standard in the Doré  context. As a reminder, the Court in  Vavilov—while expressly excepting Doré  from the scope of the comments for now (see para 55)—said that “[t]he constitutional authority to act must have determinate, defined and consistent limits, which necessitates the application of the correctness standard” ( Vavilov, at para 56).  As I have said before, this should logically include Charter issues. But Stacey does not address this point, nor does he address important literature attacking administrative constitutionalism as a general theory (see Leonid Sirota’s paper here).

Moreover, Stacey does not address other post- Vavilov commentary that could actually strengthen his point. For example, Paul Daly argues that all issues going to the merits in  Vavilov are, on its own terms, subject to the reasonableness standard. I have my issues with this argument, but I think it is far more persuasive in support of Stacey’s argument than the evidence Stacey actually offers–in part because it takes Vavilov on it’s own terms.

Relatedly, Stacey argues that a joint-reading of  Vavilov and Doré  renders the correctness standard irrelevant (349). Yet this is not convincing to me. As I have argued, and as Professor Daly argues to a similar extent,  Vavilov is not just one thing, easily explained with reference to a catchphrase like “culture of justification.” There are various currents of administrative law thought coursing through the decision.  Vavilov’s comments on the Rule of Law, for example, are relatively formalistic, focusing on the role of the courts as the guardian of the Constitution. On the other hand, other parts of  Vavilov clearly draw from the justificatory school of administrative law thought, championed by scholars like David Dyzenhaus. As I have argued in previous work, these schools of thought can be complementary, but  Vavilov is clearly a product of pragmatic agreement, even if guided by principle to some extent. Professor Stacey does not address this reality when he excludes correctness from the standard of review equation, without as much as addressing the counter-arguments clearly presented in Vavilov. Indeed, if one follows  Vavilov’s formalistic side, correctness review still has a valuable—and formally required—role in Canadian administrative law.

Finally, I should draw attention to Stacey’s argument on Charter values. Much has been written on Charter values, and I need not reprise that literature to make my objection: Stacey’s focus on Charter values as justificatory resources is only useful if the set of justificatory resources is relatively bounded and determinate. While we cannot expect perfect or near-perfect determinacy in law, and moral reasoning with regards to rights-claims is inevitable, this does not logically entail an embrace of Charter values. The problem is that Charter values are endlessly indeterminate—they are not necessarily bounded by the text of the guarantees they are supposed to represent, and some Charter values could conceivably not be found in the text. Enterprising courts and litigants could pitch a value at high level of generality, leading to needlessly subjective moralizing about rights in a way untethered to the doctrine of various constitutional guarantees. Since there is no clear agreement on (1) how to determine what Charter values are relevant; (2) how Charter values are different than Charter rights; and (3) on how administrators are supposed to understand Charter values as distinct from Charter rights, this set of justificatory resources is not at all helpful to courts or litigants.

Tying together Stacey’s article is a common claim: “…I see no distinction between administrative and constitutional law in the first place” (357). Of course, this is a common (one might say orthodox) position. And yet it ignores an important function of constitutional law in relation to the administrative state—the Constitution (written and unwritten) is a limitation or constraint on government action (see  Vavilov, at 56). The hierarchy of laws exists for a reason, and under that hierarchy administrative discretion is parasitic on a statutory grant, which itself is subject to Charter scrutiny. There is no real, formal equality between administrative law and constitutional law: the Constitution is supreme, and it shapes and constrains government power. It does not liberate administrative discretion.

All told, Professor Stacey’s article contributes to the growing post- Vavilov literature. Unfortunately, I do not find it convincing.  Nonetheless, the Supreme Court will eventually deal with Doré  post- Vavilov. And until then, my opinion is worth as much as the page it’s written on, for what it’s worth.

Overcoming Justice Abella’s Admin Law Legacy

On the occasion of her retirement, what can we learn from Justice Abella’s administrative law generation?

All good things must come to an end, and such is the case with the careers of our Supreme Court judges. On July 1, 2021, Justice Rosalie Abella will retire. Justice Abella has been a lighting rod—for good and bad reasons—throughout her tenure on the bench. There is no doubt that she, having been a Supreme Court judge since 2004, has left her mark on various areas of Canadian law. Others will analyze Justice Abella’s legacy in those areas.

In this post, I hope to provide an assessment of Justice Abella’s legacy in the world of administrative law. For a generation, Justice Abella (even before she was on the Supreme Court–see her decision in Rasanen) was a leading Canadian administrative law thinker with skills of persuasion. While Justice Abella’s thinking on administrative law was broadly representative of the judicial and academic zeitgeist of the period starting with CUPE , time and experience have shown limitations in this thinking, and the Court has rightly begun to rollback the “innovations” of this period. The problems are two-fold: (1) Justice Abella’s notion of deference is largely based on illusory assumptions about administrative expertise; (2) Justice Abella’s notion of deference makes too much of the position of administrative actors as “partners” in the law-making enterprise, especially on constitutional questions.

I will start by outlining Justice Abella’s general theory of administrative law, as represented in extrajudicial writing and some select opinions. I will then flesh out my criticisms of Justice Abella’s administrative law legacy, showing how and why the Court was justified in Vavilov in resiling from some of the commitments demonstrated by Justice Abella through her opinions and the Court’s pre-Vavilov case law. In short, Justice Abella’s lack of skepticism about government power—particularly administrative power—simply does not register as credible in the 21st century. To develop a doctrine of deference that is attuned to the diffuse nature of administrative power, the Court must continue to overcome the administrative law commitments of Justice Abella’s generation.

***

I have written many times about the dominant mode of administrative law thinking in Canada, culminating in the jurisprudential watershed moment of CUPE. Justice Abella fits neatly in this generational movement. The so-called functionalists (people like Justice Abella, John Willis, Harry Arthurs) were high on administrative power for two overlapping reasons. First, they saw the conservative common law courts stymying administrative decision-making, which was the means used by legislatures to implement social justice policy (see, particularly, the work of Harry Arthurs). Second, they assumed that administrative actors were more expert in the administrations of their statutory schemes than courts (see , again, Arthurs). This was best represented in the Supreme Court’s Dunsmuir decision, when the Court adopted the famous quote from David Mullan, suggesting that deference “recognizes the reality that, in many instances, those working day to day in the implementation of frequently complex administrative schemes have or will develop a considerable degree of expertise or field sensitivity to the imperatives and nuances of the legislative regime”: see Dunsmuir, at para 49.   The net result of these two commitments is an ardent belief that administrators can render decisions, using their own techniques, that make sense of the law—and that courts should respect those decisions. For Justice Abella, then, the “Rule of Law”—which typically justified the subjection of administrative power to the law—is a formalistic legal principle that unnecessarily pits judicial power against administrative exigency. Instead, as Justice Abella says in a co-written piece with Teagan Markin, the law should inculcate a “mutually respectful relationship between the courts and administrative decision-makers” one that prizes the “legitimacy” and “authority” of administrative actors (Abella & Markin, at 272) as a “constitutional participant” (Abella & Markin, at 298).

With these commitments in mind, the Court developed a theory of deference that did just that. In the high-water mark era for epistemic deference, the Court confirmed that expertise was the most important factor influencing the selection of the standard of review (Southam, at para 50). In Dunsmuir, as noted above, respecting the institutional choice to delegate to experts was seen as a valid reason for courts to defer to administrative actors.

Justice Abella’s functionalist mindset was clearly represented in her opinions. Three are relevant. First is her decision in Newfoundland Nurses. That case—which Vavilov implicitly overruled (see Vavilov, at para 96)—permitted courts to “supplement” decisions that were otherwise deficient in their reasons (Nfld Nurses, at para 12). What was required under Nfld Nurses was a “respectful attention to the reasons offered or which could be offered…” (Nfld Nurses, at para 11 citing Dyzenhaus). The upshot of this is that courts could not quash decisions simply on the basis of the quality of the reasons alone (Nfld Nurses, at para 14) because of considerations of specialization and expertise (Nfld Nurses, at para 13). Here, we see the translation of administrative law theory into administrative law doctrine. Because Justice Abella is concerned about the legitimacy of the administrative state and undue judicial interference, she would rather courts partner with administrative decision-makers in supplementing decisions rather than subverting them.

A second example, and perhaps the most important one, is Justice Abella’s opinion for a unanimous court in Doré. Doré is important because it demonstrates the two strands of Justice Abella’s administrative law thought: pluralism and expertise. Doré basically held that when administrative decision-makers make decisions that engage constitutional rights, their decisions are entitled to deference if they represent a proportionate balancing between the relevant Charter right and the statutory objective at play (Doré, at para 57). For Justice Abella, though, Doré was more than just a technical framework. For her, Doré was “deference theory at work” (Abella & Markin, at 299) because it showed, finally, that “administrative bodies have then authority and expertise to interpret apply…legal constraints…” (299). Even on constitutional matters, typically jealously guarded by the judiciary, this was true. Indeed, it was because of the supposed (though unproven) expertise of decision-makers on constitutional questions arising in their ambit that deference was justified (Doré, at para 47).  Here, the various strands of functionalism are in full force, yielding a rather major shift in doctrine: courts must defer to administrators on constitutional matters.

Finally, Abella and Karakatsanis JJ’s opinion in Vavilov is perfectly representative of the sort of administrative law thinking that, as I will note, may be on its way out. In Vavilov, the Court implemented a number of changes to judicial review doctrine in Canada. One of these changes was a downgrading of expertise as a reflexive or presumptive deference factor (see Vavilov, at para 30). Now, there would be no assumption that “expertise” leads to deference—expertise would need to be proven through robust reasons (Vavilov, at para 31).  Additionally, the Court also clarified that in certain circumstances, the Rule of Law—as an apparently standalone, unwritten principle— would dictate that a standard of review of correctness should apply (Vavilov, at para 53 et seq). In these regards, and as I have written in other work, Vavilov  (at least in part) represents a more formalistic template of administrative law theory than what preceded it (796). For Abella and Karakatsanis JJ, this was the problem: to them, Vavilov was a “encomium for correctness and a eulogy for deference” (Vavilov, at para 201). Why? Because gone was expertise as a presumptive reason for deference, with the substitution of a “court-centric conception of the rule of law rooted in Dicey’s 19th century philosophy” (Vavilov, at para 240).

With these three cases taken together, Justice Abella’s views on administrative law can be sketched out. She often demonstrates an abiding trust in administrative decision-makers and their expertise over legal—and even constitutional—matters. This leads to a positioning of the administrative state as a partner in law-making and interpretation. Deference, put this way, is a recognition by a judicial actor of this apparently constitutional role of administrators. While this appears to be Justice Abella’s view on administrative law, it is worth noting that the Court as a whole seemed to largely accept this understanding of the relationship between courts and administrators.

***

Having set out the gist of Justice Abella’s general conception of administrative law, I now wish to show three problems with this understanding. To the extent the Court, in Vavilov, has walked back this understanding of administrative law, it should be celebrated. And while Justice Abella’s Doré opinion remains on the books, there are good reasons to think that it, too, will and should be overturned.

  1. Expertise

The first issue with Justice Abella’s view of administrative law, demonstrated throughout her tenure, is its problematic assumption of expertise. As demonstrated through the Supreme Court’s pre-Vavilov case law, and in Justice Abella’s opinions, “expertise” was a woefully underdeveloped doctrinal concept that carried with it great power. Its invocation—especially in relation to well-established decision-makers like labour boards—ensconced those decision-makers with a juridical and psychological immunity from judicial scrutiny. But the Court never explained what “expertise” meant, how it could be recognized by courts, and why a presumption of expertise (as hardened in Edmonton East) was at all empirically justified.

In fact, many issues with expertise arise that Justice Abella and others never addressed. The first and obvious issue is that “expertise” on legal matters may simply not always exist as an empirical matter. The Vavilov majority recognized this reality when it stated that if expertise is simply assumed in all cases, it cannot be a doctrinal concept that meaningfully assists a court in determining whether a particular decision-maker is actually expert (Vavilov, at para 27 ). But more importantly, the presence of expertise is based on an empirical assumption: administrators, operating within the confines of their legal schemes, can best transfer their policy expertise to the world of interpretation; their expertise can inform their understanding of their own statutory scheme, and as a result, courts should defer. But this is based on a number of unproven assumptions: (1) that particular decision-makers have relevant policy expertise; (2) that relevant policy expertise is easily transferrable to skills required to interpret statutes; (3) that relevant policy expertise will necessarily shed light on what particular legislative terms mean. Without answers to these questions, it is simply speculation to suggest that administrators possess expertise that would assist them in interpreting the law.

The assumption is even stranger when one considers constitutional questions. If courts are to defer to administrative consideration of the Constitution, a few more assumptions need to be added to the mix. It must be assumed that relevant policy expertise=relevant legal expertise=relevant constitutional expertise. While the Constitution is law, it is a sui generis law that contains its own meanings, purposes, and interpretive techniques. Absent some compelling reason to think otherwise, it is mind-boggling to simply assume that line decision-makers will reliably and expertly contribute to the meaning of the Constitution.

Secondly, the obsession with expertise in the case law and in Justice Abella’s opinions fails to recognize the dark side of expertise. The administrative state is gargantuan, and it does not only include benevolent, public-minded people applying their “neutral” expertise in authentic ways. Expertise can also cut the other way: it can lead to a decision-maker taking a myopic view of constitutional values, or otherwise subordinating constitutional or other general legal principles to the narrow exigencies of what is required by administrative “expertise” (see for example, Kerr, at 260).  Interestingly, Kerr writes in the prison context, where there is a professional environment that systematically values control over the exercise of constitutional rights–and where concerns about assumptions of expertise are grave, indeed (see the factum of the Queen’s Prison Law Clinic in Vavilov). A lack of familiarity with constitutional norms, and a professional environment that may not inhere respect for those norms, does not inspire confidence. Indeed, the Doré framework—which places constitutional “values” and statutory objectives on the same playing field, despite the hierarchy of laws—will underpower rights because it fails to accord priority to constitutional rights over administrative objectives (see the dissent of Brown and Cote JJ in TWU, at para 206).

Finally, Justice Abella’s deployment of the “expertise” label may have led her to undervalue the importance of reasoning in determining the legality of administrative decision-makers. Doré is an example of this undervaluing. In Doré, Justice Abella did not provide any detail on the standards to be used in determining whether an administrator’s reasoning met constitutional standards. It was enough  that administrator to “balance” (whatever that means) rights and objectives; indeed, in some cases, the administrator need only be “alive” to the Charter issues (TWU, at para 56). As I will note below, this is an empty theory of deference. It tends towards abdication based on faith in expertise rather than respectful deference. Relying on expertise as a faith-based reason for deference should not exclude the requirement for proper reasoning, as Vavilov confirms.

Now, the fact that administrators may not have expertise may not be fatal for Justice Abella. In her Vavilov opinion with Karakatsanis J, the judges note that internal administrative training could be a fix (see Vavilov, at para 283) rather than authorizing “more incursions into the administrative system by generalist judges who lack the expertise necessary to implement these sensitive mandates (Vavilov, at para 283). This is a nice thought, but it is a bit like allowing the fox to guard the henhouse. There is no reason to assume, without more, that administrators will undergo training sufficient to understand the Constitution, for example. Even if there was, internal training is clearly no substitute for judicial review by generalist judges. It is the very fact that judges are generalist that makes them well-suited to ensure that general legal concerns—like the Constitution—find expression in discrete administrative regimes, with their own internal pressures. And as a matter of law, judicial review must exist. In an ideal world, we would expect administrators to structure their discretion through robust legal training, and we would expect courts to act as a backstop.

2. Pluralism

A second theme seen throughout Justice Abella’s opinions is a focus on legal pluralism. As noted above, the idea is that administrators should be seen as valid contributors to the meaning of the law and Constitution—and thus, courts should not take a supervisory or command-and-control position vis-à-vis the administrative state.

Now, it should be noted that this theme presents two distinct questions. First is whether administrators should have the power to render binding interpretations of law and the Constitution. This normative point, however interesting, is somewhat moot, in part because of the success of Justice Abella’s administrative law theory over the years. Administrators, as a matter of law, do have the power to render binding interpretations of law & the Constitution, if they are delegated the power to do so (see Martin & Conway). In my view, the ability of administrators to do so is legitimate and legal. On ordinary questions of law, the legislature has validly delegated power to administrators to decide these questions in many cases. This legislative choice must be respected absent constitutional objection. On constitutional questions, the issue is trickier, but I can certainly concede that administrators should be able to render interpretations of constitutional law as a function of their subjection to constitutional norms. In other words, if the Constitution is seen as binding on all state actors (as it should be), then it is inevitable that administrators will need to deal with the Constitution. When they do so, they are determining whether the bounds of the Constitution hem in their decision-making power. This calculation is essential if administrative actors are to be bound by the law and the Constitution.

So far as this goes, the administrative state can contribute to the meaning of law. But not too much should be made of this statement. That is because, as a matter of fact and law, administrators and judges are not on an equal playing field. Judicial review necessarily implies a relationship where one body (the court) has the authority and power to correct and surveil another body (the administrator). As a matter of law, that supervisory jurisdiction must remain (see Crevier), and it may even need to occur at a certain stringency on certain questions. As a result, there can never be a perfect equality between administrators and courts, as Justice Abella suggested.

In this way, Vavilov is a drastic improvement over what preceded it. Vavilov clearly states that administrators can and do contribute to the meaning of law, even if judicial justice does not resemble administrative justice (see Vavilov, at para 92). As far as it goes, this is an accurate descriptive statement that acknowledges the current state of Canadian administrative law. But Vavilov does not counsel abdication to administrative power. It instead insists on stringent reasoning requirements, particularly as regards the law (see Vavilov, at para 108 et seq) with only a small margin for error (Vavilov, at para 122). By doing so, it ensures that courts have standards by which they can assess administrative exercises of power, without unduly trenching on jurisdiction delegated to an actor besides the courts.

What we see here is a difference between deference as rooted in the supervisory role of the courts and deference rooted in some external appreciation of the administrative state. In our constitutional system, it is simply the reality that there must be judicial review. The way courts review administrative action puts them in a supervisory position over delegated power. This hierarchy is inescapable. Courts can–and have–developed doctrines of deference based on notions of legislative supremacy. But that doctrine of deference is quite different than one based on expertise. In the former case, deference is plausibly rooted in a exercise of constitutional power by a coordinate branch of government. Deference is not justified by a court assuming–without more–that a decision-maker could come to a “better” decision than the court. As a side note, all of this makes the last piece of Justice Abella’s administrative law legacy–Doré –vulnerable. As I wrote in this paper, the downgrading of expertise as a reflexive reason for deference and the role of the Rule of Law in anchoring the standard of review (correctness on constitutional questions) at least raise the question of Doré ‘s long-term health.

Justice Abella, in her recent co-authored article, argues that such assertions on the basis of formal constitutional materials provide no answer to her conception of administrative law. She and her co-author note that the Secession Reference, which gave a place of priority to unwritten principles of constitutional interpretation, “acknowledges the political nature of law and embraces the idea that although the government is of course constrained by legality, legality is itself a political question capable of sustaining several answers” (295). To the authors, the Secession Reference ushers in a new era that demonstrates that all institutions can take part in the making of law, lending new legitimacy to the administrative state. So, an argument as I have made on the basis of the Constitution–to the authors–is a non-starter.

Needless to say, I find this retort particularly unconvincing for a number of reasons. First, whether law is “political” or not is besides the point. While law is the product of politics, interpretive pluralism should not be taken as an excuse to simply favour the decision-makers that form the political valence we may prefer. It is the legislature’s political choices–not the court’s–that are relevant in determining the space for deference.

Secondly, Vavilov throws a ton of cold water on Justice Abella’s understanding of the Secession Reference. The Secession Reference endorses the Rule of Law as an unwritten principle of constitutional law that can give rise to substantive obligations. To the Court in its various cases, the Rule of Law is understood in a formal sense, as having to do with the subjection of government power to rules in a system of positive laws (see Secession Reference, at para 71). This is a largely formal understanding of the Rule of Law. As an analogue to this understanding, the Court has held that the Rule of Law and s.96 Constitution Act, 1867 together protect the role of the superior courts in conducting judicial review (see, again, Crevier) and protecting core superior court powers (see MacMillan Bloedel). This formal understanding of the Rule of Law was extended in Vavilov. The Court held that legislatures were not free to set up the administrative state as theu wished: legislatures could only specify the standard of review “within the limits imposed by the Rule of Law” (Vavilov, at para 35).

This understanding of the Rule of Law as an unwritten principle, and its relationship to administrative pluralism, should not be understated. Under this understanding, the Rule of Law protects not only the existence of judicial review, but it prevents legislatures from insulating administrative actors from curial scrutiny at a certain intensity on certain questions. The fact that the Rule of Law and s.96 are understood in this way serve to make a point: it would be unconstitutional, in fact, for legislatures to make administrators perfectly equal to superior courts, in a legal sense. The role of superior courts is protected constitutionally, in part, because of its importance in maintaining the Rule of Law. This invites a hierarchical relationship between courts and administrative decision-makers.

All told, the retirement of Justice Abella will be a landmark moment for the Court in many ways. And given Justice Abella’s popularity in the legal community, I have no doubt her retirement will be appropriately marked. But, as lawyers, the retirement of a prominent judge presents us an opportunity to review the body of her work. In the world of administrative law, Vavilov represents the first major effort to overcome Justice Abella’s persuasive legacy. This is welcome.

No matter what, I wish Justice Abella well on her retirement.

An Oddity in Strom

In October, the Saskatchewan Court of Appeal released its much-anticipated decision in Strom. Strom raised a number of important issues: “ “at the intersection between professional regulation, Ms. Strom’s private life, and the s.2(b) Charter guarantee of freedom of expression in the age of social media.”  

Strom was a registered nurse. Her grandfather tragically passed at a long-term care facility. Strom took to Facebook to criticize the care her grandfather received at the facility. The facility’s employees reported the comments to the Saskatchewan Registered Nurses’ Association (SRNA). The SRNA charged Strom with professional misconduct, and the SRNA Discipline Committee found her guilty.

The Court ultimately overturned the Discipline Committee’s decision. For many reasons that I cannot explore here, I think this is the right decision, in law and in principle. But one aspect of the decision is of particular interest to me: the Court’s standard of review discussion as it related to the freedom of expression arguments raised by Strom.

Of course, in such a case, the framework that governs the standard of review analysis is Doré. Doré holds that the standard of review when a court reviews an administrative decision is reasonableness (Doré, at para 7). Doré also introduces a proportionality framework for assessing whether a decision-maker has struck a reasonable balance between the Charter right at hand and the statutory objective. Post-Vavilov, it is at least an open question about whether Doré is still good law. This is because Vavilov reaffirms that, when litigants challenge a law under the Constitution, the standard of review is correctness (Vavilov, at para 56). I, for one, have questioned why it is that different standards of review should apply, especially since the Court in Vavilov recognized that  legislatures cannot “…alter the constitutional limits of executive power by delegating authority to an administrative body” (Vavilov, at para 56).

Enter Strom.  In that case, both parties agreed that the standard of review is correctness on the Charter issue (Strom, at para 133). The Court also agreed, but only because the case came to court via a statutory right of appeal, and under Vavilov, statutory rights of appeal invite the appellate standards of review (correctness on questions of law, palpable and overriding error on questions of fact/mixed fact and law) (see Strom, at para 133). The Court noted, though, the following, at para 133:

It is not necessary to consider the question left unanswered by Vavilov, at paragraph 57; that is, what is the standard of review when the issue of whether an administrative decision has unjustifiably limited Charter rights is raised on judicial review, rather than on appeal?

This, in itself, is not really problematic. Given the fact that, for now, Doré lives another day, it would not be appropriate for an appellate court to apply the correctness standard to Doré-type situations. It is defensible—and proper—to simply classify a constitutional issue as a question of law that falls under the scope of an appeal right. And at the end of the day, it does not matter much for a results perspective, because the standard will be correctness either way.

However, in choosing the correctness standard, the Court then did something that is difficult to understand. It said the following, at para 140:

What, then, is an appellate court’s task when reviewing whether the decision of an administrative body unjustifiably infringed a Charter right? In substance, that task is summarily described in Doré at paragraph 6, despite the fact that the standard of review is correctness. The Court’s task is to determine whether the decision-maker disproportionately limited the Charter right or struck an appropriate balance between the Charter right and statutory objectives.

The Court went on to describe the question raised by the case, at para 166:

The question as to whether it has imposed excessive limits is the proportionality question. Here, it is whether the Discipline Committee advanced its statutory objective in a manner that is proportionate to the impact on Ms. Strom’s right to freedom of expression. One aspect of that question is whether the impact on her freedom of speech in her private life was minimal or serious.

The Court then went on to apply the Doré framework, though made reference to some parts of the Oakes test (see para 153).

This strikes me as an oddity. Let’s take what the Court describes its task to be. When courts apply the correctness standard, courts focus on “the conclusion the court itself would have reached in the administrative decision maker’s place” (Vavilov, at para 15). The proportionality analysis advanced in Doré, however, seems to have deference built-in to it. While Doré notes that the Oakes test and the Doré framework “exercise the same justificatory muscles,” (Doré, at para 5), Doré deference asks courts to give some weight to the statutory objective being advanced by the decision-maker. This was, indeed, a bone of contention for the dissent of Brown and Côté JJ in Trinity Western. But when a court conducts correctness review, at least in theory, the court should not give any weight to what the decision-maker’s reasons are for making

That said, there are no perfect Platonic forms in law. It is true that Oakes itself has developed to contemplate deference in its application. And it is also true that courts, post-Vavilov, have applied what I call “light correctness review” (see Planet Energy, at para 31), where a certain amount of weight is given to the decision-maker’s “…interpretation respecting the words of the Act, the general scheme of the Act and the policy objectives behind the provision.” This could be seen as a sort of Skidmore-like deference, under which courts give non-binding weight to an administrative interpretation.

That said, the inherently deferential idea of Doré review seems inconsistent with a stringent application of the correctness standard. This, to me, is a theoretical oddity, even if its effect is blunted on the edges.

On John Willis and the Pesky Politics of Administrative Law

John Willis was and is considered one of Canada’s most important administrative law academics. As a student of administrative law and the law of judicial review, one cannot skip Willis’ classic works, like his books “The Parliamentary Powers of English Government Departments” and “Canadian Boards at Work”—and his caustic papers, including his attack on the McRuer Report and his famous “Three Approaches to Administrative Law: The Judicial, The Conceptual, and the Functional.” Especially in this latter piece, Willis sets out his comprehensive functional theory of the law of judicial review—that theory holds that courts, with a typically “conservative” orientation, could never understand the difficulties of governments that, post-New Deal, were concerned with social welfare. Accordingly, courts should butt out, in service of the expertise, efficiency, and progressive orientation of administrative decision-makers.

While Willis should obviously be commended and respected as an eminent scholar of administrative law, there is a core problem at the centre of Willis’ thought that should stand as a warning for us in the modern era. Willis fundamentally viewed administrative law as a project of politics. As R. Blake Brown notes in this article, and as Willis himself always argued, the law of judicial review (and administrative law more broadly) was not about legal principles or controls on the administrative state, but was rather designed to limit the interference that legalist, conservative courts could wrought on tribunals pursuing the social good in an expert way. But this sort of thinking runs into two fundamental problems: (1) it ignores the fact that, strategically, administrative power can and has been used to fulfill the policy aims of governments who do not have any designs on social welfare goals—this was a clear blind-spot in Willis’ own work, one that led him to over-trust government; and (2) normatively, as recent Supreme Court and Federal Court of Appeal jurisprudence tells us, there is a meaningful difference between law and politics when it comes to the interpretation and application of laws governing judicial review. Granting deference based on the supposed underlying political motivations of particular decision-makers undermines this separation.

Let’s start by reviewing Willis’ functionalism. Willis self-described himself as a “government man,” and “what actually happens man.” By this he meant he was less concerned about the legal principles of a 19th century constitution, but rather was concerned with the making of “effective government” (see “Administrative Law in Retrospect” at 227). Functionalism crafted doctrine to align with the way government operates and the programs government is responsible for implementing. At the time of Willis’ writing, the struggle for government was the implementation of social welfare programs, closely associated with the New Deal. Delegation to administrative tribunals was one of the ways that these programs were implemented. A functional approach, then, would respect the legislative choice to delegate

In my view, deference to administrative tribunals in the functionalist mould was supported by a number of presuppositions about administrative decision-making, but the most important one for our purposes was what I call the “political” presupposition. Deference on the functionalist account was justified because of the apparent political valence of the decision-makers under review and what they were designed to achieve. Says Michael Taggart (at 257), describing the functionalists of the era:

These left-leaning scholars were deeply resentful of what they saw as conservative judges twisting the pliable rules of statutory interpretation to favour the existing order, privileging the rich and the powerful, and defeating the purposes of statutes intended to further the interests of the workers, the homeless, and the least well-off in society.

Deference on functionalist grounds was therefore a reactionary force, one that was a political project designed to fight back against the supposedly conservative orientation of the courts, that used legal principles to stunt the social welfare programs of governments.

Willis himself clearly fell in this category. A social democrat, Willis railed against any sort of thinking that would interfere with the prerogatives of government, undertaking social programs. He viewed government as fundamentally changing in light of the New Deal:

The State had changed its character, had ceased to be soldier and policeman, and was rapidly becoming protector and nurse…Again the right of the community bulk larger than the rights of the individual (See Parliamentary Powers at 13 and 51).

Risk described Willis’ functionalism as such:

Willis’ thinking about law and government can best be summarized by dividing it into three parts. The first is his observations and attitudes about government and its institutions….He perceive the nature and extent of the expansion of government, and its implications for the structure and functions of the legal institutions. He perceived a changing relation between the individual and the community, and how legislative policies were expanding the claims of the community against the individual, and circumscribing common law ideals….He had a great faith in experts, and he believed the courts should give liberal scope to the agencies on review (see Risk, at 545).

The political appreciation of administrative tribunals as representing the needs and wishes of the “community” was a constant thread through Willis’ scholarship advocating for a judicial “hands-off” approach to decision-makers like labour boards.

So, what are the problems with such an approach? To my mind, there are two. First, Willis’ political approach to deference fails on its own terms: it fails, on any complete account, to actually reach an ideal of social justice. This is because deference itself has no political valence, and can be easily used to vindicate decisions of administrators that run counter to social justice. And secondly, on principle, a political approach to deference runs counter to our positive law and to the good reasons for it.

Let’s first tackle the issue of social justice and its connection to doctrines of deference. As I’ve written time and time again, using deference as a means to reward the  decision-makers we like –because we ascribe to them some political ideal—is an unprincipled and politically naïve way to view the law. Let’s start with the latter contention. Willis’ supposition seems to be that courts themselves cannot be trusted to uphold the purposes of ameliorative legislation because of their conservative orientation. But it is not axiomatically true—and in fact, it seems bizarre—to suggest that deference will always serve to advance social welfare principles. It does no such thing.

The development of the administrative state is simple proof of this. When Willis was writing, he made the near-sighted appreciation that deference supported the administrative state as it existed at that time. At that time, observers were mostly concerned with labour tribunals, who were seen as consummate experts in their craft. But Willis either did not predict that deference would and could also have to attach to tribunals he did not appreciate under his social welfare rubric. That is, and I have said before, there are other aspects of the administrative state that do not map so neatly onto any past or modern description of social justice. Prisons, an area of interest for me, come to mind—perhaps the place where administrative discretion, at least prior to the CCRA, was most unbounded. Another example, that of immigration decision-making,  also comes to mind. What, beyond brash politics, justifies treating these tribunals any differently?  If one believes in deference, how can one say that prisons are any less deserving of deference than any other decision-maker?

This does not strike me as a consistent approach based on social justice. Later in Willis’ career, this thought must have occurred to him. That is because, in his “Administrative Law in Retrospect,” Willis addressed the question of a number “fashionable cults” which, to him, were negatively affecting the prerogatives of government:

This is very interesting coming from someone who is committed to social justice. How could it be consistent, if one accepts some political justification for deference based on social justice, for Willis to deny prisoners a right to be heard? Why are prisoners less good, in Willis’ eyes, than unions before a labour board? Willis’ myopic conception of social justice was profoundly underinclusive, even on its own terms.

Another explanation of this oddity is that Willis was not committed to social justice at all. Rather, it is very likely and possible that Willis was indeed a “government man” in the most literal sense of the term. Any action that could offend a government prerogative, in his eyes, was abhorrent. So the “cult of the individual,” and prisoners, all serve to run against government, even if government offends social justice. But this stands inconsistently with the idea of social justice. Social justice, on any cohesive account, is not about empowering government for the sake of government. The problem is that government can act in ways that contradict the theory of social justice

The above point challenges Willis on his own social justice terms, but there is an external, doctrinal reason to be wary of Willis’ approach to deference. It is indeed true that the fight for deference in Canada is overlayed by considerations of politics. After all, the laws delegating power to decision-makers—or laws that work to limit the scope of power for these decision-makers—are passed against the backdrop of a legislature that is a partisan organ. But that is a separate matter altogether from the actual legal justifications for deference, which like the interpretation of statutes, should be a separate concern from politics. Luckily, our law recognizes this fact clearly, and does so for good reasons. Vavilov, for example, does not base deference on any good-faith presumptions about the expertise of decision-makers. Now, the very fact that the legislature delegated power—any power—to any decision-maker is a fact worthy of deference. The tool of delegation as the grounds of deference has the benefit of being agnostic as to how one can judge, politically, particular tribunals. And Vavilov itself (at paras 120-121) cautions against reverse-engineering doctrine to suit a desired outcome. Clearly, Willis’ political approach to doctrine (and the arguably political approach of the common law courts) run afoul of our current law, which erects a clear separation between deference as a doctrine and the political results of a deferential approach. Justice Stratas in the Federal Court of Appeal has made a similar point in the context of statutory interpretation and judicial review: see Williams, at para 48; Cheema, at para 74; and Hillier, at para 33.

While that is the state of the positive law, it is the positive law for a good normative reason. It is orthodox today in the academy that law cannot be meaningfully separated from politics. It is even true that some say that any attempt to do so is necessarily “reactionary” or “conservative.” But this contention does not take account of the different parts of law-making and interpretation, and the very purpose of law itself. As I mentioned above, it is of course true that laws reflect the political consensus of the legislature at the time they are passed. It would be wrong and overbroad to suggest that the making of law is or should be divorced from the political process: indeed, it is the function of our legislatures to make laws that, at least in theory, are undergirded by the support of a majority. However, this is a completely separate act from the interpretation of laws. Laws, in order to be consistent with at least one aspect of the Rule of Law, must be general rather than specific; and when a judge interprets a law, she does so to give the meaning to the text, context, and purpose of the law that is enacted on the page (even purpose, as I discuss here, is usually and ultimately guided by text). The task of interpretation of laws should not be governed by consideration of politics; of what this or that judge thinks of this or that tribunal. Should it be the case that judges grant deference because of their political views, we will go along way towards undermining our separation of powers between courts and legislatures, imperfect as it is in Canada.

Of course, it is impossible to say that politics will never enter the interpretive activity. But that is a different question altogether than how doctrine is constructed. Ideally, the way we theorize deference and interpretation should not be based on political musings; rather, theory should be based on the foundational principles of our legal order, including the choice of a legislature to delegate power and the core interpretive function of the courts. It might be orthodox to suggest otherwise these days, but in my view, the very purpose of law in the law of judicial review is to enforce the limits that legislatures themselves provide—no matter how wide or loose they are—on administrative decision-makers. It should not be the role of the courts, as Justice Stratas so eloquently says in the above-cited decisions, for courts to pick winners and losers based on politics.

Back to Willis. John Willis’ contributions to Canadian administrative law will live for the ages. But his approach to the law of judicial review should not be celebrated wholesale. Willis’ cardinal mistake was falling victim to the game started by the “conservative” common law courts. If it is true that those courts struck administrative decisions because they did not appreciate the social welfare function of those agencies, that runs counter to our governing law and the good reasons for it. But today, Willis is still celebrated; the common law courts are not. I think it is fair and appropriate to draw attention to the blindspots in Willis’ theory: his myopia regarding what he thought was “social justice”; and the specious attempt to import deference based on some political justification.

Vavilov in the Prisons

By now, Vavilov—the case in which the Supreme Court re-jigged Canada’s standard of review framework—has received sustained attention, including from yours truly. Over at Administrative Law Matters, Professor Daly has a running post outlining how Vavilov has been applied in some particular interesting cases. And on SSRN, Jamie Chai Yun Liew has an excellent article on the implications of Vavilov for marginalized persons.  Vavilov has had and will continue to have implications for particular areas of law.

My concern in this post is the way Vavilov is being applied, at least in a few cases, in an important area of administrative decision-making: the carceral state. Indeed, it is not often that prisons are mentioned as distinct areas of administrative decision-making. This is, perhaps, because the administrative state is often identified closely with the “welfare state”—a benevolent set of tribunals and decision-makers maximizing benefits for vulnerable people. This is best represented in the deification of labour tribunals as the quintessential administrative decision-maker: expert, oriented towards the public good, efficient, and cheap.

But the prisons are not some separate and apart actor in terms of administrative law. Prison decision-makers operate on delegated power, just as labour and human rights tribunals do. Prisons have become increasingly bureaucratized over time (see Malcolm Feeley and Edward Rubin’s classic text here), and with that bureaucratization has come the potential for the traditional pitfalls associated with administrative decision-making. Indeed, as I will point out in the context of a particular case, “expertise” in the prison context means something very different—it often means expertise in using punitive measures to “manage” “difficult” prisoners.  Coupled with the intersecting disadvantages of prisoners, and the level of control prison administrators—presumably without legal training—hold over those same prisoners, the risk of unlawful or unconstitutional decisions affecting the statutory or constitutional rights of prisoners multiplies. Indeed, prisons were once referred to as “lawless agencies” for good reason—they form, like immigration decision-makers, an intimate part of the administrative state as we know it.

Hence it is important to study the way in which judicial review doctrine works in the carceral state. Indeed, one would expect that Vavilov’s focus on justification might actually make a difference in the prisons, where decisions have historically been made without the close scrutiny of courts. The following two cases illustrate the divergent ways in which, at least in these two cases, Vavilov is faring against the behemothic carceral state. In one of the cases, the court takes a deferential posture based on “expertise,” arguably undermining the justificatory promise of Vavilov. In the other, the court subjects the prison decision to the scrutiny required by Vavilov’s justificatory standard.

Bromby v Warden of William Head Institution, 2020 BCSC 1119

Bromby is serving an indeterminate life sentence for first-degree murder. He was involuntarily transferred from a minimum security institution to a medium security institution in 2019. He objected to this transfer, “[r]aising concerns about the sufficiency of disclosure” in relation to the transfer [2]. Despite this objection, the transfer was approved by the Warden of the minimum security facility. However, Bromby raised a habeas corpus claim, and was successful on that claim. He was transferred back to the minimum security institution.

However, immediately following this decision, the minimum security facility initiated an “emergency involuntary transfer” based on an increase of his security classification score. The final decision to transfer Bromby from minimum to medium security was eventually made by the warden, and was delivered orally to Bromby—but he was not provided written reasons for this decision  in a timely manner, contrary to the Corrections and Conditional Release Regulations [4]. Bromby argued that “the transfer on an emergency basis was unreasonable as he did not pose a threat to the security of the penitentiary or the safety of the inmates or any other person” [5].

In response to this claim, the Warden trotted out an old standard of prison decision-making: the rather specious appeal to “micromanagement” and “expertise.” As the respondent submitted:

  1.    It is not the role of reviewing courts to micromanage prisons even where they feel that intervention measures other than a transfer might have been taken in addressing inmate behaviour. While the applicant may have preferred for other actions to be taken to attempt to manage his risk, deference is owed to the Warden’s assessment that the applicant was unmanageable within Mission (Minimum) Institution.

. . .

  1.    The Decision was . . . reasonable . . . based on the facts and legal constraints before the Warden. The decision of the Warden, a prison administrator familiar with the complexities of Mission (Minimum) Institution and the security classification of inmates, should be provided with deference to decide that the applicant was presently incapable of management within an open perimeter environment. The thorough explanations provided and thoughtful insight as to the specific interventions that the applicant can work towards in becoming a minimum security inmate reflect the Warden’s significant expertise in identifying and managing offender risk. Accordingly, a high degree of deference is owed to the Warden in his decision.

 

The Court largely accepted this line of thinking:

However, it is the January 2020 Decision that I must assess for reasonableness and determine whether it falls within the range of possible acceptable outcomes which are defensible on the facts and the law. I find that the decision does. That is because the warden is entitled to deference in the decision‑making process. The decision of the warden, a prison administrator familiar with the complexities of Mission Institution and the security classification of inmates, set out the basis for why it was that Mr. Bromby presented as being incapable of management within an open‑perimeter environment [63].

There are two general problems with this line of thinking.

First, I confess that I don’t understand how this line of thinking is at all consistent with Vavilov. Vavilov did away the presumption of expertise for administrative decision-making, instead focusing on “demonstrated expertise” (Vavilov, at para 93). It is true that this demonstrated expertise “may reveal to a reviewing court that an  outcome that might be puzzling or counterintuitive on its face nevertheless accords with the purposes and practical realities of the relevant administrative regime…” (Vavilov, at para 93). This does, fairly, give some latitude for prison decision-makers such as the Warden in this case to apply their knowledge of a particular situation to a dispute. But in this case there is no interrogation of the demonstrated expertise of the Warden. That is, the court does not determine whether the supposed expertise of the Warden was actually demonstrated in the reasons. The fact that, in the court’s view, “[t]he record establishes that the warden turned his mind to all of the relevant considerations…” [65] is not enough to warrant an acceptance of pre-Vavilov law on expertise. This has particular resonance in the prison context, where expertise has often been assumed without demonstration (see Lisa Kerr’s wonderful article here).

I acknowledge that it is genuinely difficult to demonstrate, in the prison context, what actually constitutes “expertise.” Vavilov opens the door to the operational realities of prisons—including issues of security—factoring into a decision. But there is no critical assessment here by the court of how the Warden’s expertise featured into this decision. This seems to be what Vavilov prescribes, and it arguably should factor into any assessment of reasonableness, particularly where the consequences are dire for the individual (Vavilov, at para 133).

Moreover, it is important to recognize that “expertise,” as a general proposition, and especially in the carceral state, can refer to many different things. Simply stating, as the court does in Bromby, that wardens have expertise masks the real question: in what? As Lisa Kerr notes in another outstanding article, at 259, expertise can cut both ways, especially where constitutional rights are at stake. Expertise could be a veneer for stereotyping or discrimination, as I note below.  This is an important normative reason to deny administrative decision-makers a presumption of expertise, which Vavilov explicitly rejects—but which has, based on Bromby, perhaps not filtered down to the carceral state.

Finally, I must acknowledge the old trope about “micromanagement” of prisons, trotted out by the respondent in Bromby. Judicial review is, it is true, not about micromanagement of administrative decision-making. It is about policing the boundaries of statutory schemes designed to cabin administrative activity; it is a traditionally legal and doctrinal activity. As Kerr again aptly acknowledges:

The organization dynamics of prisons tend to resist constitutional constraints, due to the political powerlessness of inmates and the structural isolation of corrections from the community. The status of the inmate is defined in relation to managerial goals, rather than in relation to an externally defined moral norm, and prison managers tend to focus on their vision of scientific management rather than the larger legal order. Amid these institutional tendencies, only the judiciary has the inclination and ability to impose a regular and comprehensive legal framework. The judiciary is a necessary play in prison legality, rather than a necessarily amateur outsider at risk of “micromanagement.” The spirit of habeas corpus, with its strict emphasis on legality and access to justice, so as to challenge deprivations imposed on the physical body, has always had this in mind.

This is fundamentally important. Judicial review is a check against the seductive force of administrative exigency, in which people might be assimilated based on stereotypes or useful organizing tools rather than as individuals. Yet prisons, at the same time, must acknowledge the rights (statutory and constitutional) of prisoners—this is acknowledged in CSC’s enabling legislation.  Ensuring that these rights are upheld is the function of judicial review, which should not be reduced to some afterthought when evaluating the panoply of control mechanisms at the hands of prison decision-makers.

Johnston v Canada (Atorney General), 2020 FC 352

Contrast the previous case with Johnston out of the Federal Court. Johnston involved an inmate at Kent Institution. As a federal inmate, Johnston received payments from Correctional Service Canada “at a modest daily rate” [1]. However, CSC began withholding 100% of Johnston’s modest pay. This is because he had not paid a costs order in favour of the Attorney General. As such, Johnston filed a judicial review of the decision to make the a 100% reduction in his inmate payments.

The Court concluded that, though the relevant statutory scheme gave the CSC the legal authority to make deductions, “it was unreasonable for the CSC to withhold all of the applicant’s income without considering the purpose and principles that govern CSC and without considering the impact the deductions would have on the applicant” [4].

The problem in this case was marred by issues with the record. As the Court noted, the only record of decision was an email chain originating in the CSC. That email chain revealed that there was no “explanation or justification for making deductions from the applicant’s inmate income” apart from the obvious costs order [14]. Specifically, the record did not say “anything about…why the amount of the deduction was set at 100 percent of the applicant’s inmate income” [15].

While Vavilov does note that reasons are not required in every case (Vavilov, at para 77), reflecting well-known law, the case does note that “where reasons are provided but they fail to provide a transparent and intelligible justification…the decision will be unreasonable” (Vavilov, at para 136). This is precisely what happened here. In addition to failing to disclose why the 100% figure was chosen, the CSC failed to consider the vulnerability of the person who had has income reduced. This is directly contrary to Vavilov (see Vavilov, at paras 133 et seq).  The Vavilov majority puts the point eloquently at para 135:

[135]                     Many administrative decision makers are entrusted with an extraordinary degree of power over the lives of ordinary people, including the most vulnerable among us. The corollary to that power is a heightened responsibility on the part of administrative decision makers to ensure that their reasons demonstrate that they have considered the consequences of a decision and that those consequences are justified in light of the facts and law.

The CSC’s conduct in this regard was sorely lacking in terms of the justificatory standard set out in Vavilov—probably more so than Bromby. When a court cannot glean a reasonable justification from the record and reasons—in this case, a generally worded email—the risk increases that the administrative action was arbitrary. The risk increases that, in this particular case, the number of 100% was plucked from the air, without any discernible reason. When we multiply this arbitrariness with the existing vulnerability of prisoners, we have a recipe for administrative disaster.

The point is not that prisoners, because of their vulnerability, must win every judicial review. Doctrine must be applied without fear or favour. But the doctrine includes the consideration of the circumstances of vulnerable persons, and the importance of a decision to those persons, raising the justificatory bar in those cases (see, for this point, Sharif, at para 9). Again, this is not a trump card, but it is an important consideration for administrative decision-makers. It is not something to discard in favour of administrative exigency.

Justification plays a useful role here. It forces the prison, which is by design isolated from the rest of the community, to articulate the reasons for decisions in a way that is understandable to the people within prisons, as well as to the external legal system. Justification is the window by which we can look into the morass of prison decisions, policies, and directives that bear on the actual lives of real people within the carceral state. The tendency in prisons is, likely, to reject these external checks.

That makes those checks all the more important.

The Continued Relevance of “Jurisdiction”

This post is co-written with Leonid Sirota

One of the innovations of Vavilov was its dispatch of so-called “jurisdictional questions” from the standard of review analysis. A long-time feature of Canadian administrative law, jurisdictional questions were said to arise “where the tribunal must explicitly determine whether its statutory grant of power gives it authority to decide a particular matter” (see Vavilov, at para 65; Dunsmuir, at para 59). These questions would attract correctness review. But as the Vavilov majority acknowledged, “…majorities of this Court have questioned the necessity of this category, struggled to articulate its scope and ‘expressed serious reservations about whether such questions can be distinguished as a separate category of questions of law” (Vavilov, at para 65; Alberta Teachers, at para 34).   As a result, the Court decided that it would “cease to recognize jurisdictional questions as a distinct category attracting correctness review” (Vavilov, at para 65), satisfied in the knowledge that the robust reasonableness review it articulated would solve a potential problem of decision-makers arrogating power to themselves they were never intended to have (Vavilov, at para 68; para 109).

We question whether matters are so simple. While the Court purported to rid Canadian administrative law of “jurisdictional questions,” clearly the concept of jurisdiction remains. In this post, we outline the four ways in which it remains relevant in Canadian administrative law, despite its absence from the standard of review analysis. This happens (1) in the course of statutory interpretation under Vavilov itself; (2) in the presence of certain statutory rights of appeal; (3) when drawing the boundaries between the remits of two or more tribunals; and (4) when determining whether a tribunal is empowered to consider Charter questions.

A note before beginning: between us, we view questions of jurisdiction differently. One of us (Mancini) has previously argued that jurisdictional questions should simply attract reasonableness review, since jurisdictional questions are merely a subset of a larger category of questions of law; in his view, there is no meaningful difference between jurisdictional questions and other questions of law, for the purposes of the standard of review (see the reasons of Stratas JA in Access Copyright (2018) at para 75). The other (Sirota) disagrees with this position, and instead believes that questions of jurisdiction must attract a correctness standard of review, and that if this means that most or all questions of law, being jurisdictional in some sense, require correctness review, so much the better. This difference is not material for the purposes of this post. We only mean to argue that the Vavilov judgment should not be read as dispensing with the existence of all questions of jurisdiction, let alone with the concept of jurisdiction writ large. Indeed, jurisdiction still remains an important and relevant concept in distinct areas of Canadian administrative law, an idea recognized in some respects by Vavilov itself.

Statutory interpretation under Vavilov

As noted above, Vavilov ceases to recognize jurisdictional questions as a distinct category attracting correctness review (Vavilov, at para 65). This is not a surprise, as majorities of the Court had previously thrown doubt on both the concept of jurisdiction (see CHRC, at para 38) and the means used to identify jurisdictional questions (McLean, at para 25).

And yet: chassez le naturel, et il revient au galop. When the Court goes on to describe the statutory context within which a particular decision-maker operates as an “obvious and necessary constraint” on administrative power (Vavilov, at para 109),  the Court’s explanation harkens back to the language of jurisdiction. The Court says that

Reasonableness review does not allow administrative decision-makers to arrogate powers to themselves that they were never intended to have, and an administrative body cannot exercise authority which was not delegated to it (Vavilov, at para 109, our emphasis).

What is this if not an invocation of the concept of jurisdiction, albeit in plain English? Whether we frame the issue as one of statutory authority or jurisdiction, the point is the same: administrative decision-makers only have the power that is explicitly or impliedly delegated to them by legislation (or that they hold under the royal prerogative). If they go beyond the scope of the delegation, the decision-makers lose their authority to act. Far from doing away with the concept of jurisdiction, then, the Court embraces it in its articulation of the legal limits of reasonableness review.

Moreover, the Court explains that “[i]f a legislature wishes to precisely circumscribe an administrative decision maker’s power in some respect, it can do so by using precise and narrow language and delineating the power in detail, thereby tightly constraining the decision maker’s ability to interpret the provision” (Vavilov, at para 110). In such cases, “questions relating to the scope of a decision maker’s authority … may support only one” permissible interpretation (Vavilov, at para 110), by contrast with others where the statutory language is more open-ended. While the Court resists the analogy, it is difficult to distinguish single-answer “questions relating to the scope of a decision maker’s authority” from pre-Vavilov questions of jurisdiction. What is new, perhaps, is the implicit recognition that even open-ended grants of authority must have their limits.

This is not something to be worried about―even though, as the Vavilov majority noted, every question regarding an administrative decision-maker’s statutory limits can be conceived as a question of jurisdiction (see Vavilov, at para 66), and is so conceived elsewhere (see Peters v Davison (NZCA) explaining that UK case law, followed in New Zealand, has served to “render redundant any distinction between jurisdictional and non-jurisdictional error of law”). Indeed, the Court is correct in saying that jurisdiction (or statutory authority) is a natural limit on administrative discretion. Although it does not serve as the lynchpin for a distinct category of legal questions for the purposes of standard of review analysis, the concept remains in the articulation of the limits on administrative decisions.

Statutory Rights of Appeal and Privative Clauses

Under Vavilov, different standards of review apply on statutory appeals and on judicial review. On appeal, when a case involves a question of law, the standard will be correctness; when a case involves a question of fact or mixed fact and law, the standard will be palpable and overriding error. On judicial review, by contrast, most questions of law, as well as questions of fact and policy, attract reasonableness review.

Hence the scope of statutory rights of appeal, and thus whether a given issue can be appealed or must be judicially reviewed, may be decisive for the outcome of a case. This scope can be circumscribed; one common way in which this is done is by limiting the right of appeal to “questions of law and jurisdiction” as, for example, in the Broadcasting Act provision at issue in Vavilov’s companion case, Bell/NFL.

How are such provisions to be interpreted? Vavilov could be read in one of two ways on this score. First, one could read Vavilov to suggest that when a legislature provides an appeal on a question of law or jurisdiction, jurisdiction means the same thing as “law.” This appears to be what the Court did in Bell, when it did not mention the difference in legislative language between questions of law or jurisdiction. Secondly, one could read Vavilov as retaining the concept of jurisdiction, but simply concluding that for standard of review purposes, the distinction between law and jurisdiction does not matter. This retains the concept of jurisdictional questions.

But what if the appeal right only extends to questions of jurisdiction, not to non-jurisdictional questions of law?  (See, for a version of this in Quebec, Mancini’s article on the subject). If this happens, there are three options. If Vavilov is read as saying that the concept of jurisdiction has no distinct meaning, courts can safely ignore the privative clause and simply consider the right of appeal as either extending to questions of law, or perhaps as covering a null set of cases. We find either of these solutions to be undesirable. If a legislature uses the term “jurisdiction” in a right of appeal, in contrast to the term “law” in a privative clause, the legislature’s use of that term must be given effect: this is simply an application of the presumption against tautology, endorsed in Vavilov itself (see para 45). If the legislature uses the term jurisdiction in a statutory right of appeal, it must mean something over and above a question of law, however much courts and scholars might disagree with its implicit determination that there exist non-jurisdictional questions of law.

This means that courts, in determining whether a particular matter falls within such a right of appeal, must come to its own determination about whether the subject matter is “jurisdictional.” Jurisdiction, then, continues to rear its head in these scenarios.

Jurisdictional Boundaries Between Two or More Administrative Bodies

The Vavilov majority retained, as a category of question attracting correctness review, the determination of “jurisdictional boundaries between two or more administrative bodies” (Vavilov, at para 53).  This happens when it is unclear which body must resolve a given issue, and one body attempts to address a matter that may be considered to fall within a comprehensive legislative regime administered by another.

The Court says that “[a]dministrative decisions are rarely contested on this basis” (Vavilov, at para 64). This observation is true, but the category is not without controversy. In fact, the Court will hear a case, Horrocks, which considers the demarcation of the respective spheres of authority of human rights tribunals and labour arbitrators, and the governing test for determining which actor should assume jurisdiction in a particular case (see Weber, Figliola). In these cases, the Court uses “jurisdiction” in its standard sense: as the power to hear and decide certain matters. If a tribunal proceeds erroneously on this score, it would incorrectly assume jurisdiction.

It might seem puzzling that Vavilov retained this category of review while purporting to rid Canadian administrative law of other “jurisdictional questions.” And yet, what choice did the Court have? As it pointed out, litigants (and indeed tribunals themselves) need to know which administrative body is tasked with resolving a given question.

Jurisdiction to Consider Charter Questions

The question of whether a decision-maker can consider the Charter is also a question of jurisdiction in the classic sense. It is noteworthy that the term “jurisdiction” appears 89 times in the Supreme Court’s reasons in Martin, which set out to re-appraise the rules governing whether a decision-maker has the authority to consider Charter issues. This is a preliminary question that must be asked before dealing with the merits of a particular constitutional challenge. The Court in Martin concluded that where there is jurisdiction to decide questions of law, there is also jurisdiction to consider the Charter (see Martin, at para 36). For the Martin Court, jurisdiction is defined as “the power to decide” (Martin, at para 36). It will be a “jurisdictional question,” therefore, whether a decision-maker has power to determine how the Charter applies to a matter on which it is required to rule. When a court reviews a decision-maker’s conclusion on this front, the court will owe the decision-maker no deference (see Martin, at para 31).  In this manner, the concept of jurisdiction will continue to inform whether a decision-maker has power to decide a Charter matter, and such questions will function much the same way as they did pre-Vavilov.

This isn’t to say that this category of review is justified from a perspective of first principles or precedent. The Constitution is always a limitation on government action, whether that action is legislative or administrative. That is, legislatures should not be able to “delegate out” of the Constitution by empowering an administrative actor. While it is true that administrative decision-makers are creatures of statute, constitutional constraints circumscribe statutory grants of authority whether they are mentioned or not. Indeed, the better view is that a legislature cannot preclude a decision-maker from considering the Constitution even by saying so. And from the perspective of precedent, Martin is difficult to reconcile with Doré, which held that “…administrative decisions are always required to consider fundamental values” (Doré, at para 35). While we both consider Doré to be unjustified in every other respect, this aspect of Doré―at least if for the extra-constitutional “values” we substitute the more appropriate “law”―is supported by the fundamental idea that the Constitution is supreme in the hierarchy of laws: s.52 of the Constitution Act, 1982 (see also Canada (Citizenship and Immigration) v Tennant, 2018 FCA 132).

Normative Implications

In our view, the holding in Vavilov on jurisdictional questions must be considered quite limited. The Court must not be taken as saying that “jurisdictional questions” do not exist as a conceptual matter. Nor is the Court saying that, in other contexts, courts must defer on questions that can be identified as jurisdictional.

Rather, the situation is much more nuanced. Jurisdiction remains a relevant principle in Canadian administrative law, in at least four areas where courts will be called upon to delineate the scope of the authority of particular decision-makers, whether in the ordinary process of statutory interpretation, in demarcating jurisdictional lines, construing statutory rights of appeal, and in relation to Charter questions. Courts will need to return to a stable definition of jurisdiction. It will do no good to suggest that “jurisdictional questions” have been banned from the world of Canadian administrative law. Horrocks is an example: there, the Court will need to decide whether its test for determining which particular body has jurisdiction is adequate.

In our view, this narrow reading of Vavilov is normatively desirable. Jurisdiction is not the will-o’-the-wisp some make it out to be. Scholars obsessed with the “bad old days” of pre-CUPE administrative law always speak of jurisdiction as if it is some major impediment to administrative decision-making. But that is only so if administrators must, contrary to basic constitutional principles requiring all public power to be constrained by law, be allowed to roam free of legal fetters. Such claims by the defenders of the administrative state are an admission against interest, and quite an unnecessary one. Administrative decision-makers function just fine in jurisdictions where their jurisdiction and, indeed, the correctness of their legal interpretations are fully policed by the courts.

It is true that judges of a particular era were pre-disposed to view administrative power with skepticism. But they had good reason: the rise of administrative power was not an inevitability or a phenomenon that was totally consistent with fundamental constitutional principles. Jurisdiction—the idea that a law (typically statute but sometimes the common law) that exists outside the administrator’s subjective preferences and is subject to judicial interpretation determines whether the administrator can hear or decide a matter—is merely a constitutionally required limit on administrative power (see Vavilov, at para 109). No amount of tinkering with standards of review can change this. Courts trying to flee from constitutional principles will find that they cannot outrun them. They must reckon with this reality and devote their energy to working out how these principles are to be applied, rather than to futile escapades.