Public Law’s Cerberus: A Three-Headed Approach to Charter Rights-Limiting Administrative Decisions

Abstract

This article offers a theoretical and doctrinal solution to a vexing question in public law: how to determine the justifiability of Charter rights-limiting administrative decisions. The jurisprudence suggests three approaches, or modes of reasoning: minimal impairment analysis, ‘interest balancing’, and ‘values-advancing reasoning’. Like Cerberus, the guard dog of Hades, Canadian public law has become three-headed. While scholars and courts argue about which mode of reasoning is categorically best, the culture of justification compels us to ask instead which provides the most compelling explanation for each rights-limiting decision. Just as cutting off one of Cerberus’s heads would diminish his effectiveness as a guard dog, rejecting either of the modes of reasoning would limit decision makers’ capacity to explain their decisions and undermine a culture of justification. The article makes a theoretical case for retaining all three modes of reasoning and sets out a doctrinal approach to determining when each is applicable.

1. Introduction

With the 2012 decision in Doré v Barreau du Québec, the Supreme Court of Canada affirmed that the principle of proportionality is central to Canadian public law. ¹ That much is hardly controversial. Since at least 1986, statutory limitations on Charter rights have been assessed for their proportionality through the four-step formula set out in R v Oakes. ² The thing about Doré that has stoked debate, however, is the novel analytical approach the unanimous Court instructed administrative officials to take when making decisions affecting Charter rights.

Administrative decision makers, the Court said, perform different tasks to legislatures and in different contexts. The assessment of whether an administrative decision has unjustifiably limited Charter rights should therefore differ from the assessment of whether a statutory provision does so. ³ Accordingly, administrative decision makers and in turn reviewing courts need not proceed through all four stages of the Oakes formula when considering whether an administrative decision limits a Charter right. Rather, decision makers should focus their analytical attention on only the fourth element of Oakes, the inquiry into proportionality ‘in the strict sense,’ and ask whether their decision strikes a proportionate balance between the rights limitations and the statutory objectives achieved by those limitations.

Doré has not been well received. ⁴ Even the Supreme Court has not been consistent in its application of the Doré approach in subsequent cases. ⁵ Much of the criticism suggests that Charter rights will be better protected if administrative decision makers and courts of review rely instead on the Oakes test in full. Doré has come to be vilified as some kind of terrifying rights-destroying monster, like Cerberus, the three-headed dog of Ancient Greek mythology who confines human souls to Hades. ⁶ And just as Hercules tamed Cerberus, Doré’s critics call for its vanquishing by brave and heroic jurisprudence.

I argue in this paper that demonizing Doré is misguided. The criticisms levelled at it are logically flawed, and the doctrinal approaches its critics have proposed are unhelpful. My objective, then, is to offer a different doctrinal solution to a messy area of public law. I suggest that the minimal impairment inquiry integral to the Oakes formula and the ‘balancing’ approach set out in Doré are merely different heuristics, or modes of reasoning, by which administrative decision makers can explain their decisions. At the core of my argument is the culture of justification to which the Supreme Court unambiguously and emphatically committed in its 2019 decision in Vavilov. ⁷ It is clear after Vavilov that administrative decision makers (almost always) have an obligation to explain their decisions to the people they affect. ⁸ This is no less the case when those decisions impose limits on Charter rights.

I argue here that neither the Oakes formula nor the Doré approach is categorically better at explaining why a rights limitation is justified. On the contrary, the logic on which each approach works is suited to providing compelling explanations for rights-limiting administrative decisions in different circumstances. Administrative decision makers and courts of review can foster a culture of justification by relying on whichever mode of reasoning provides the most accessible and intelligible explanation in the circumstances. The Doré approach is ultimately just as committed to justification as the Oakes approach, and the difference between them lies more in form than in substance.

A close reading of Doré and the cases that followed reveals three distinct modes of reasoning about the justifiability of rights limitations. The first is faithful to the Oakes formula and relies heavily on the inquiry into minimal impairment or the availability of less restrictive means. It does not rely on balancing or proportionality ‘in the strict sense.’

Two other modes of reasoning, in contrast, foreground proportionality in the strict sense. The second mode involves a decision maker weighing up the costs and benefits of rights limitations directly against each other in determining if a limitation is justified. This mode of rights reasoning has been called “interest balancing.” ⁹

I call the third mode of reasoning ‘values-advancing reasoning’. This approach allows a decision maker to decide whether to limit a right in pursuit of statutory objectives or uphold the right and forego the statutory objectives, depending on which of those options better advances Charter values.

Reasoning about rights limitations in the administrative context is thus, not unlike Cerberus, a three-headed creature. But like Cerberus, I argue that Doré has been misunderstood. Cerberus is no more than a dog with an important job to do, namely, to maintain the balance between the Underworld and the living by ensuring the dead do not escape Hades. Cerberus has three heads to allow him to do that job effectively: all three of his heads are dedicated to his task and not one of them is categorically better at it than the others. Cerberus is no monster, even though he is three-headed. His role in the ecosystem of Greek mythology is critical.

Reasoning about rights limitations is the same: three-headed, unwieldy, and perhaps scary to some, but nonetheless critical to Canadian public law. Like Cerberus’ heads, each of the three modes of reasoning about rights-limiting administrative decisions is committed to the culture of justification. None of them should be ignored or severed, and in suggesting that we should rely only on Oakes analysis in all cases, Doré critics risk undermining the culture of justification. Retaining all three modes of reasoning ensures that administrative decision makers will be in the best position to explain their decisions whatever circumstances they find themselves in.

And after all, Hercules did not vanquish Cerberus. He captured him, presented him to the Mycenaean king Eurystheus as one of twelve labours he was required to complete, and then returned him to his post at the gates of the Underworld, no worse for wear. Even Hercules, that great slayer of monsters and beasts, recognised the critical role that Cerberus plays in the balance between life and death.

I argue that Doré, too, is an important statement about how to maintain the integrity of Canadian public law. ¹⁰ Reading Doré in the context of Oakes and alongside subsequent cases, we can see how it provides decision makers and judges with a range of tools on which to rely in inquiring into the justifiability of rights limitations and finding balance between Charter rights and broader policy objectives.

I present the argument as follows: In Part 2, I revisit the Doré decision, focussing on the reasons the Court gave for departing from the Oakes formula. I admit upfront that Doré is a vexing decision. Among other things, it leaves two questions unanswered: how are decision makers actually supposed to proceed with the inquiry into the justifiability of a rights-limiting decision; and what role do Charter values play in the proportionality analysis? I try to fill these voids in Parts 3 and 4, and describe interest-balancing and values-advancing reasoning in more detail as I do so.

In Part 3, I consider how two subsequent Supreme Court decisions adjust and refine the Doré approach, laying the groundwork for the interest balancing mode of reasoning. ¹¹ In Part 4, building on the Court’s decision in Groia v Law Society of Upper Canada, I explain the role of Charter values in the proportionality analysis and outline the third mode of reasoning: values-advancing reasoning. ¹²

I draw the discussion to a close in Part 5, bringing legal theory and case law to bear in arguing that less restrictive means analysis and the two proportionality-centric modes of reasoning are each valuable, in different circumstances. I engage with Doré’s critics here and reject their insistence that Doré’s proportionality-centric approach is somehow less effective in protecting rights than Oakes’ focus on less restrictive means analysis. I end with the important conclusion that having three modes of reasoning provides decision makers with heuristic options by which they can ensure, in every situation, that Vavilov’s demands of justification are met.

2. Proportionality Analysis in Administrative Decision Making

On its face, Doré is an important judgment that changes public law analysis in Canada, just as other judgments have before it. The list of doctrine-altering landmark cases will be nauseatingly familiar to any upper-year law student or practitioner: Nicholson, CUPE, Pushpanathan, Baker, Dunsmuir, and of course Vavilov. ¹³

These judgments have, variously, affirmed that administrative decisions can be reviewed for reasonableness (CUPE); developed an ever-sharpening conception of reasonableness and outlined increasingly workable approaches to reasonableness review (Vavilov); declared that everyone must be treated fairly, not arbitrarily, regardless of the form of the administrative decision (Nicholson); identified considerations going to what would be procedurally fair in the circumstances (Baker); and set out a variety of tests for determining which standard of review—reasonableness or correctness—is to apply (first Pushpanathan, then Dunsmuir, and now Vavilov).

In many ways, these landmark cases have moved Canadian public law away from a rigid, formalised, rules-based model that relies on categorical reasoning and box-checking, towards a more flexible approach oriented to upholding fundamental principles. Central to this move has been a deepening commitment to the culture of justification, culminating in the ‘robust’ conception of administrative reasonableness the Supreme Court set out in Vavilov.

But in other ways, the case law of the Supreme Court has retained a commitment to formalism. Vavilov itself commits to formalistic reasoning in the box-checking version of the standard of review analysis it introduces. Gone is the contextual inquiry adopted in Dunsmuir, replaced by an inquiry that asks only whether the decision falls into certain categories.

As a member of this club of landmark cases, Doré is notable for its decidedly lukewarm reception. Much of the criticism it has faced, though, reflects the formalism tenaciously embedded in Canadian public law. Vavilov’s two-facedness in this regard threatens Doré’s legacy even at the same time that it vindicates the basic logic on which Doré rests. ¹⁴

Reluctance to embrace the Doré approach stems from the view that focusing on the reasonableness of the balance that an administrative decision maker strikes between rights and statutory objectives leaves Charter rights open to erosion at the administrative level, and at the mercy of less exacting review at the judicial level. By proceeding formalistically through all four stages of the Oakes analysis, the argument goes on, decision makers and review courts will be better able to ensure that administrative decisions correctly uphold Charter rights. Engaging with these critics’ arguments requires, as a first step, a review of the approaches set out in Oakes and Doré.

The Oakes test is a familiar one to most public lawyers, in Canada as well as around the world. Many legal systems commit to the principle of proportionality in rights analysis, and versions of the Oakes test exist in many of these countries. ¹⁵ Just like these other tests, the Oakes test provides a formulaic method by which courts and legislatures can think through whether a statutory rights limitation meets the requirements of reasonableness and demonstrable justifiability set out in section 1 of the Charter. In all, the Oakes test includes four distinct inquiries. The limitation of a Charter right is only justifiable if all four of the inquiries are satisfied.

First, the objective a rights-limiting statute seeks to achieve must be sufficiently important to justify limiting rights. The limitation must have a “pressing and substantial” objective if it is to be constitutionally acceptable. ¹⁶ Second, the limitation must be rationally connected to its objective. If the statutory provision is unlikely to achieve its goals anyway, the rights limitations it imposes cannot be justified. Third, the limitation must be minimal: it must restrict or impair Charter rights no more than is reasonably necessary to achieve its objectives. The conclusion that there are less restrictive means available by which the legislature could achieve its pressing and substantial goals suggests that the limitation is more damaging to rights than it needs to be and is therefore disproportionate.

Finally, even if the limitation does or is likely to achieve pressing and substantial goals and impairs rights as little as possible in doing so, the impairment must nevertheless be proportionate to the benefits it achieves. This inquiry into proportionality in the strict sense asks, bluntly, whether the benefits the limitation delivers are worth the harm it causes. For the limitation of a right to be justifiable, it must strike a balance between harms and benefits.

As it turns out, the third leg of the analysis—the inquiry into whether the limitation is minimally impairing or whether there are less restrictive alternatives—has done most of the work in Canada. There are relatively few cases where the Supreme Court has struck down a rights limitation on the basis that it is minimally impairing but nonetheless disproportionate in the strict sense, and only one I can think of in which the Court upheld a rights limitation as proportionate in the strict sense even though less restrictive alternatives were open to the legislature. ¹⁷

In most cases, the Court has let its conclusions on the availability of less restrictive means guide its inquiry into proportionality in the strict sense. If less restrictive means were available to the legislature to achieve its objectives—that is, if the rights limitation was not minimally impairing—the statutory measure is usually struck down on that basis alone: the courts rarely go on to even consider whether the limitation is proportionate in the strict sense. ¹⁸ And where the Court finds that a limitation is minimally impairing—that is, where there are no reasonably less restrictive alternatives available—the inquiry into proportionality in the strict sense is usually brief and perfunctory.

The Court’s record in this regard has led commentators to conclude that proportionality in the strict sense is never determinative, by itself, of the justifiability of a rights limitation. ¹⁹ In their comprehensive work on Canadian constitutional law, Peter Hogg and Wade Wright argue that the final step of the Oakes analysis duplicates the other steps in the analysis, “has no work to do, and can safely be ignored.” ²⁰ Even members of the Supreme Court have suggested that the inquiry into proportionality in the strict sense merely rehearses the considerations in earlier steps of the formula. ²¹

It is thus hardly controversial that the Supreme Court relies overwhelmingly on the less restrictive means analysis in determining the justifiability of statutory rights limitations, and that proportionality in the strict sense carries little of the analytical burden. It must have come as something of a surprise to many public lawyers, then, when in Doré the Supreme Court instructed administrative decision makers to skip the first three inquiries of the Oakes analysis, including the load-bearing less restrictive means inquiry, and ask only whether the decision strikes a proportionate balance between statutory objectives and rights. Indeed, only six years earlier, a majority of the Court had confirmed that Oakes provided the appropriate analytical framework for assessing the justifiability of rights-limiting administrative decisions. ²²

Abella J’s unanimous judgment in Doré gives three reasons for refocusing the analysis on proportionality in the strict sense. The first is that there are significant differences between administrative officials applying legal rules to specific people in specific cases, and legislatures drafting statutes that are intended to apply generally across society. The Court began,

In assessing whether a law violates the Charter, we are balancing the government’s pressing and substantial objectives against the extent to which they interfere with the Charter right at issue. If the law interferes with the right no more than is reasonably necessary to achieve the objectives, it will be found to be proportionate, and, therefore, a reasonable limit under s. 1. ²³

The Court then contrasted the law-making context with the administrative decision making context: “In assessing whether an adjudicated decision violates the Charter, however, we are engaged in balancing somewhat different but related considerations, namely, has the decision-maker disproportionately, and therefore unreasonably, limited a Charter right.” ²⁴

The Court emphasised later on in the judgment that it is difficult to conduct the first three inquiries of the Oakes test “where there is no specific enactment that can be examined in terms of objective, rational connection, least drastic means and proportionate effect.” ²⁵ Indeed, the Court had already departed from a strict application of the Oakes framework in other situations where statutory provisions are not involved, such as in a challenge to the constitutionality of a common-law rule. ²⁶

The second reason the Court gave for abandoning the first three inquiries of the Oakes test is that administrative law theory had advanced significantly since the Court first held that the Oakes framework was appropriate in the context of rights-limiting administrative decisions. It may have been true in 1989 that the judicial review of administrative decisions for reasonableness rested “to a large extent on unarticulated and undeveloped values and lack[ed] the same degree of structure and sophistication of analysis” as the proportionality test set out in Oakes. ²⁷ But by 2012, especially after Dunsmuir had consolidated patent unreasonableness and reasonableness simpliciter into a single standard of review in 2008, the Court apparently believed that a richer, more robust and more sophisticated conception of administrative reasonableness had emerged. ²⁸

Central to the Court’s reasoning here is how a focus on balancing rather than proceeding through all the stages of the Oakes formula can ensure that officials exercise their discretion “in light of constitutional guarantees and the values they reflect.” ²⁹ It is simply unnecessary to retreat to a full Oakes analysis if decision makers “are always required to consider fundamental values.” ³⁰

The third reason the Court gave for the revised approach builds on this point: decision makers have a great deal of experience and expertise in their own areas of decision making. As a result, they are likely not only to be able to identify the relevant Charter values at stake, but also to have insight into how those values are best advanced in particular situations. ³¹

Towards the end of the judgment, the Court set out what we might think of as the Doré approach. If an administrative decision maker need no longer rely on all four stages of the Oakes formula in considering whether a decision imposes justifiable limits on Charter rights, how are they to proceed?

He or she balances the Charter values with the statutory objectives. In effecting this balancing, the decision-maker should first consider the statutory objectives…. Then the decision-maker should ask how the Charter value at issue will best be protected in view of the statutory objectives. This is at the core of the proportionality exercise, and requires the decision-maker to balance the severity of the interference of the Charter protection with the statutory objectives. ³²

In Vavilov, the Supreme Court declined the invitation to revisit Doré. ³³ As a matter of precedent, then, Doré continues to set out the approach for administrative decision makers to follow when making rights-limiting decisions. But precedent is a merely formal reason to uphold a judgment. The core of my argument here is rather that the robust approaches to substantive review that Doré and Vavilov describe, for decisions affecting Charter rights and decisions attracting reasonableness review respectively, share a normative foundation.

To insist on Oakes’ formulaic and formalistic approach to decisions affecting Charter rights, as Vavilov’s formalistic elements might compel us to do, would undermine Vavilov’s own normative roots in a culture of justification. Doré, for its part, provides an approach to rights-affecting administrative decision making that is of a piece with the normative commitment Vavilov makes to a culture of justification. ³⁴ This normative consistency is the reason we should be sceptical of any claim that Vavilov provides reasons to overturn Doré.

However, as a blueprint for how to go about determining whether or not a decision justifiably limits Charter rights, the handful of lines the Court offered at the end of Doré are not all that helpful. The case law that followed Doré highlights at least two matters on which the Court’s reasons in that judgment do not provide a great deal of guidance.

The first is what it means for a decision maker to balance Charter protections and statutory objectives, and how decision makers are supposed to go about this now without relying on any of the other inquiries of the Oakes heuristic. I take up this question in Part 3, below. In Part 4 I turn to the second area of confusion, which concerns what a Charter value is and how it differs from a Charter right. As I answer these questions in Parts 3 and 4 of the paper, I outline the second and third modes of reasoning that decision makers can rely on, beyond Oakes and minimal impairment analysis, to determine whether rights-limiting administrative decisions are justifiable or not.

3. How to Doré: Interest Balancing Rather Than Minimal Impairment

a) The first restatement: Loyola High School v Quebec

The first Supreme Court case after Doré to deal with a rights-limiting administrative decision was Loyola High School v Quebec. ³⁵ Loyola is a Catholic high school that had applied to the Quebec Minister of Education, Recreation and Sports for an exemption from the requirement to offer a mandatory non-denominational program in ‘Ethics and Religious Culture’(ERC). ³⁶ Loyola had proposed an alternative to the ERC program, which the statutory framework authorised the Minister to approve if it was ‘equivalent’ to the ERC. The Minister did not find Loyola’s proposed alternative to be equivalent, however, concluding that it was faith-based rather than secular and amounted to a departure from the objectives of the ERC. ³⁷ Loyola sought judicial review on the basis that the decision unjustifiably infringed the right to freedom of religion.

Abella J wrote the majority judgment, with McLachlin CJ and Moldaver J writing a minority judgment concurring in part. Having begun the judgment with a reminder that the Court had “eschew[ed]” the Oakes formula in favour of a “robust proportionality analysis consistent with administrative law principles,” the majority’s analysis unsurprisingly focused on the balance the Minister struck between the ERC program’s objectives and the right to freedom of religion. ³⁸

The majority noted that the ERC, for its part, was intended to foster diversity and inculcate respect for all religions and peoples. The program explored the elements of religion without taking a denominational or dogmatic angle. It encouraged students to reflect on the values and norms central to different religious groups and to interact respectfully with members of other faiths. ³⁹

The Minister’s explanation for refusing to exempt Loyola’s alternative program was that it did not take a neutral perspective in offering instruction in the Catholic religion. In both the majority and the minority’s view, this was not a justifiable limit on Loyola’s religious freedom. For Abella J and the majority, requiring the school “to teach Catholicism from a neutral perspective” imposed a heavy burden on freedom of religion with little advancement of the ERC program’s statutory objectives. ⁴⁰ Moreover, the Minister’s reasoning did not give any weight to the implications that her decision held for religious freedom: “There is, in short, no balancing of freedom of religion in relation to the statutory objectives.” ⁴¹

The majority did uphold the part of the Minister’s decision that insisted that Loyola teach non-Catholic spiritual worldviews from a neutral perspective. To allow Loyola to teach other ethical frameworks through the lens of Catholic dogma would run the risk of belittling non-Catholic belief systems. This, the majority said, would undermine the very objectives of the ERC program. On the other hand, requiring Loyola’s teachers to offer instruction on non-Catholic religious worldviews from a neutral perspective imposes only minor limitations on the right to freedom of religion. ⁴² In this respect, the Minister’s decision did strike an appropriate balance between the right to freedom of religion and the program’s objectives.

McLachlin CJ and Moldaver J appear to take a markedly different analytical approach than the majority, even though the judgments are not that different in the result. Most glaring in the minority’s reasons for judgment is the almost complete failure to mention Doré. The minority refers to Doré once in its judgment, and only then in its summary of the matter’s procedural history. ⁴³ The minority does not look to Doré for guidance in assessing the justifiability of the Minister’s decision.

On the contrary, the minority hews quite closely to the approach set out in Oakes, assessing the justifiability of the limitation by asking whether the Minister’s decision limited the right as little as reasonably possible. Finding that the Minister understood the ‘equivalency’ requirement of the exemption regulation to require schools to propose alternative programs that essentially replicated the content of the ERC program, the minority concluded that the Minister not only rendered nugatory the purposes of the exemption but limited the right to religious freedom extensively. A more flexible interpretation of the regulation, allowing more deviation from the ERC program, would still align with the objectives of the program without interfering with rights as substantially. “In short, the Minister’s decision was not minimally impairing.” ⁴⁴

While the minority is quite clear in the extent to which it relies on the minimal impairment or less restrictive means inquiry, the majority’s approach—ostensibly an application of the Doré approach—is riddled with less restrictive means analysis. Consider the majority’s gloss on Doré in the following paragraphs:

Under Doré, where a discretionary administrative decision engages the protections enumerated by the Charter … the discretionary decision-maker is required to proportionately balance the Charter protections to ensure that they are limited no more than is necessary given the applicable statutory objectives that he or she is obliged to pursue. ⁴⁵

A proportionate balancing is one that gives effect, as fully as possible to the Charter protections at stake given the particular statutory mandate.… A Doré proportionality analysis finds analytical harmony with the final stages of the Oakes framework used to assess the reasonableness of a limit on a Charter right under s 1: minimal impairment and balancing.… Both [Oakes] and Doré require that Charter protections are affected as little as reasonably possible in light of the state’s particular objectives. ⁴⁶

The majority’s comments here are confusing, following so soon after the Court’s unanimous decision in Doré. They suggest that minimal impairment does have a role to play in assessing the justifiability of rights-limiting administrative decisions after all, and that the inquiry into minimal impairment and proportionality in the strict sense are related rather than separate.

I suggest in Part 5 that this is the correct understanding of the relationship between minimal impairment and proportionality in the strict sense, even if the majority judgment in Loyola and the unanimous judgment in Doré do not explain that relationship satisfactorily. I argue below that the inquiry into minimal impairment, by itself, provides an answer to the question of a limitation’s justifiability only under one condition. Spoiler alert: that condition is when the less restrictive alternatives are just as effective in achieving the pressing and substantial goals as the impugned, more restrictive option. In all other cases, where the less restrictive alternatives are also less effective than the more restrictive options, the analysis depends on one of the proportionality-centric modes of reasoning.

Even if I am right about the integrated and interdependent nature of minimal impairment and proportionality in the strict sense as a matter of theory, it does nothing to clarify what Doré and Loyola want administrative decision makers to actually do as a matter of doctrine. Are they to rely on the balancing test alone? Or proceed through the last two stages of Oakes? Or conduct a new kind of balancing inquiry that somehow blends both of those inquiries?

There is some guidance to be found, mercifully, in a close reading of the first of two Supreme Court decisions from 2018: Law Society of British Columbia v Trinity Western University. ⁴⁷ This case illustrates how interest balancing works.

b) Interest balancing in Trinity Western University v Law Society of British Columbia

As in Loyola, the litigation between Trinity Western University (TWU) and the Law Society of British Columbia (LSBC) concerned a conflict between the freedom of religion and an administrative decision maker’s exercise of discretion in pursuing statutory objectives. TWU is a private evangelical Christian post-secondary institution that has wanted to open a law school for many years. In order to enrol in the LSBC’s bar admission program, applicants must show proof of academic qualification. Graduates of law schools that are approved by the LSBC are automatically taken to have met this requirement.

TWU therefore applied to the LSBC for approval of its proposed law school. After some prevarications and reversals, the LSBC ultimately decided not to approve TWU’s law school. The main reason for this decision was that, at the time, TWU required all students to agree to a code of conduct (the “Community Covenant Agreement”) that, among other things, prohibits “sexual intimacy that violates the sacredness of marriage between a man and a woman.” ⁴⁸

In this regard, TWU had effectively adopted an admissions policy that discriminated against members of the LGBTQ community. Approving a law school that discriminates in this fashion, the LSBC reasoned, would undermine the statutory objectives it is mandated to pursue: i.e., upholding and protecting the public interest by preserving and protecting the rights and interests of all persons, and ensuring the integrity of lawyers. ⁴⁹

A majority of the Court upheld the LSBC’s decision as a reasonable exercise of administrative discretion, finding that it struck a proportionate balance between statutory objectives and Charter rights. McLachlin CJ concurred in separate reasons. Rowe J concurred in the result, finding there to be no limitation of religious freedom. Brown and Coté JJ dissented. I focus here on the approach taken in the majority judgment and on some of McLachlin CJ’s comments.

In contrast to Loyola, the majority does not appear to rely heavily on the inquiry into less restrictive means or minimal impairment. Rather, it goes directly to the balancing inquiry and considers whether the extent of the limitation on TWU’s religious freedom rights is justified by the benefits that refusing to approve TWU’s law school will deliver.

In performing this balancing analysis, the majority concluded that the LSBC’s decision did not limit TWU religious freedom to a significant extent. First, the LSBC did not refuse flat-out to approve TWU’s proposed law school. Rather, it rejected the proposal for a law school that included a discriminatory admissions policy. It remained open to TWU to make the Community Covenant optional for students rather than mandatory, which would soften a great deal of the discriminatory impact of TWU’s commitment to community values. ⁵⁰ More to the point, adherence to the Covenant is not absolutely required by the faith to which the TWU community subscribes. ⁵¹

Second, studying law in an environment where the community’s religious beliefs are pervasive, as they would be where all students are required to conduct themselves in accordance with the Covenant, was preferable to members of the community, but not necessary for their spiritual growth. ⁵²

On the other hand, the majority found that the refusal to approve TWU’s law school, at least as long as it insisted on a mandatory Covenant, was aligned with the LSBC’s statutory objectives in two important ways. First, it maintained “equal access to and diversity in the legal profession.” ⁵³ Because members of the LGBTQ community are unlikely to want to attend an institution hostile to their identity, the mandatory Covenant effectively restricted admission to the law school to persons who do not identify as LGBTQ. ⁵⁴

Second, if members of the LGBTQ community did decide to attend TWU and to agree to the terms of the Covenant, this would come at considerable personal cost: “Attending TWU’s law school would mean that LGBTQ students would have to deny a crucial component of their identity in the most private and personal of spaces for three years in order to receive a legal education.” ⁵⁵

Refusing to accredit TWU’s law school furthered the LSBC’s statutory mandate to preserve rights and freedoms, the majority held, by preventing the risk of serious harm to those members of the LGBTQ community who might have chosen to attend TWU.

In the majority’s analysis, the harms of approving TWU’s law school were significant, and outweighed the relatively minor limitations on the TWU community’s religious freedom rights. Put differently, the benefits gained by not approving TWU’s law school (i.e., promoting equal access to law schools and reducing the risk of harm to members of a minority community) far outweighed the deleterious consequences of doing so (i.e., minor limitations on the right to religious freedom). This is a prime example of the mode of reasoning that has been called ‘interest balancing’. ⁵⁶

McLachlin CJ also engaged in interest-balancing reasoning in her judgment, although for her the analysis was slightly different. She found on the one hand that the limitation of the TWU community’s religious freedom rights was serious. On the other hand, for the LSBC to have approved TWU’s proposal would have amounted to an organ of state condoning discrimination against members of the LGBTQ community. There is great value, the Chief Justice went on, in organs of state sending the unambiguous message that the Canadian government will never condone private discrimination. The LSBC’s refusal to accredit TWU’s law school sent this message, McLachlin CJ concluded, and the benefits of sending this message were great enough in her view to justify even a serious limitation of TWU’s religious freedom rights. ⁵⁷

Both the majority and McLachlin CJ adopted interest balancing as the mode of reasoning in Trinity Western. By considering whether the benefits achieved by limiting a right are greater than those that would be achieved by upholding the right, decision makers and courts of review can come to a conclusion about which course of action is constitutionally appropriate. This is an example of proportionality analysis in the strict sense, or balancing as the final stage of the Oakes analysis envisages it. Indeed, in its early section 1 jurisprudence, the Supreme Court indicated that a proper application of the final stage of the Oakes analysis would involve a contextualised weighing up of salutary benefits and deleterious consequences, rather than a comparative assessment of the abstract value of Charter rights and broader social goals. ⁵⁸

The most relevant aspect of interest balancing as a mode of reasoning—and indeed the way the majority and McLachlin CJ applied it here—is that it does not concern itself at all with whether there are less restrictive alternatives. The conclusion stands on its own logic. In this regard, note that both the majority and McLachlin CJ acknowledged that the less restrictive means inquiry was of no assistance in this case. As the majority put it, “the LSBC was faced with only two options—to approve or reject TWU’s proposed law school.” ⁵⁹ One of these options was obviously less restrictive than the other, but approving TWU’s law school only because it would have been less restrictive to do so misses the point entirely. The real question is not which option is less restrictive, but whether the option that does limit rights but achieves statutory objectives is justified.

In other words, the answer to the minimal impairment stage of the Oakes test in these situations will always be that there are, in fact, no less restrictive means by which the decision maker can pursue the relevant statutory objectives. It is worth quoting McLachlin CJ at length on this point:

I agree with the majority that on judicial review of a rights-infringing administrative decision, the analysis usually comes down to proportionality, and particularly the final stage of weighing the benefit achieved by the infringing decision against its negative impact on the right.… Minimal impairment—whether the administrative decision infringes a Charter right more than necessary or is broader than reasonably required—arises, but the question is not whether “the law” catches more conduct than it should, as under Oakes, but whether an alternative less-infringing decision was possible. Particularly where the decision is a choice between only two options (for example, to accredit or not), this step will also easily be met. This leaves the final stage of the proportionality inquiry—assessing the actual impact of the decision. It follows that in reviewing administrative decisions, the analysis almost invariably comes down to looking at the effects of the decision and asking whether the negative impact on the right imposed by the decision is proportionate to its objective. ⁶⁰

What the Trinity Western judgments leave us with is the germ of an idea that an inquiry into the availability of less restrictive means is just not useful, at least in some situations. And in situations where this question does not lead us any closer to an answer, decision makers and courts must rely on a different mode of reasoning. This is not to say that the minimal impairment test is universally useless. Rather, it is to suggest that in some situations it will not help a decision maker to determine whether or explain why a rights-limiting decision is justified or not.

I take up this idea in Part 5 below, where I sketch out the theory and the doctrine of a three-headed approach to administrative decision making where Charter rights are engaged, suggesting more precisely when minimal impairment analysis provides a determinative answer to the question of a limitation’s justifiability. Before doing that, I return to the second area of confusion that Doré begat: the relationship between Charter rights and Charter values. Bringing clarity to this relationship is relevant to developing the third mode of analysis I present here.

4. The Role of Values in Proportionality Analysis

One of the criticisms levelled against Doré (I consider other criticisms in Part 5, below) is that it requires judges and decision makers to think about the vague and imprecise concept of Charter values when balancing rights and statutory objectives. Part of the impetus for this criticism comes from the Court’s admittedly imprecise use of the term in Doré and subsequent cases, and its failure to draw a clear distinction between ‘Charter values’ and ‘Charter rights’. In Doré, the Court uses the two terms almost interchangeably, and in Trinity Western the majority judgment confoundingly talks about rights and values as a single category it calls “Charter protections.” ⁶¹

Some members of the Supreme Court have indicated their own concerns with this aspect of Doré. As both McLachlin CJ and Côté and Brown JJ (dissenting) point out in Trinity Western, whatever Charter values might be, they are not the same thing as Charter rights. Lumping the two together is jurisprudentially inaccurate and doctrinally confusing. “Charter values may play a role in defining the scope of rights,” McLachlin CJ noted in Trinity Western, but “it is the right itself … that receives protection under the Charter.” ⁶²

Côté and Brown JJ were even more direct:

Charter “values”—unlike Charter rights, which are the product of constitutional settlement—are unsourced. They are, therefore, entirely the product of the idiosyncrasies of the judicial mind that pronounces them to be so.… Lacking the doctrinal structure which courts have carefully crafted over the past 35 years to give substantive meaning to Charter rights (including the right to equality) and to guide their application, Charter values like “equality”, “justice”, and “dignity” become mere rhetorical devices by which courts can give priority to particular moral judgments, under the guise of undefined “values”, over other values and over Charter rights themselves. ⁶³

I agree that it is important to separate Charter rights from Charter values. The latter are not justiciable in the way that rights are. ⁶⁴ But whether or not values are justiciable, the fear that Côté and Brown JJ expressed in Trinity Western was that if broadly and abstractly stated Charter values are weighed against Charter rights in the balancing analysis, a wider swathe of rights limitations would be found to be justifiable. That would open Charter rights up to extensive erosion.

This fear is not unfounded, but it is overblown to the extent that, as I note above in discussing interest balancing as a mode of reasoning, the Court has long mandated that the proportionality analysis should not consider the value of rights and competing objectives in the abstract but in the specific context of each case. It is unhelpful to consider whether less hate speech is more or less valuable than freer speech, because both are valuable in the abstract. Where a statute limits certain kinds of speech, what matters is whether the actual consequences of those specific limits are proportionate to the actual benefits those limitations achieve or are likely to achieve.

But if the way to avoid this criticism is to move away from abstract values and into contextualised benefits and harms, what work is there left for broad and fundamental Charter values to do in the analysis? I suggest that there are two distinct ways that Charter values inform the proportionality analysis. The first is central to interest balancing as a mode of reasoning, and the second provides the foundation for my account of values-advancing reasoning.

a) Values in interest balancing

The Court has already articulated at least one task for Charter values, which is to guide the assessment of how serious or extensive a rights limitation is. Having a sense of the seriousness or extent of a rights infringement is critical to interest balancing as a mode of reasoning about the justifiability of rights limitations, because we have to know how serious a rights limitation is if we are to weigh its adverse consequences against the benefits it produces. This is especially useful in those situations where less restrictive means analysis turns out to be of little assistance to a decision maker.

In R v Keegstra, a case involving restrictions on hate speech, the Court noted that the seriousness of a rights limitation depends on how closely the prohibited conduct is connected to the core of a right. ⁶⁵ Elaborating on this idea in RJR-MacDonald v Canada, the Court explained that each right serves a set of “even more fundamental values.” ⁶⁶ When an activity or form of conduct does not promote those values, it cannot be said to lie close to the core of the relevant right. ⁶⁷ Prohibiting or restricting such conduct or activities will thus result in rights limitations that are not significant.

Although Keegstra and RJR-MacDonald dealt with legislative limits to Charter rights, Charter values do the same work in the context of administrative rights limitations. The majority put the point quite clearly in Trinity Western, tracing a line of reasoning through its administrative law jurisprudence back to at least the 2009 decision in Hutterian Brethren: ⁶⁸

In Loyola, this Court explained that under the Doré framework, Charter values are “those values that underpin each right and give it meaning” and which “help determine the extent of any given infringement in the particular administrative context and, correlatively, when limitations on that right are proportionate in light of the applicable statutory objectives.” ⁶⁹

Charter values, on this approach, are neither goods to be protected in the same way as Charter rights nor goods to be weighed against Charter rights in the balancing analysis. Rather, Charter values inform and sustain the content and scope of rights. In the first place, by identifying the broader and more fundamental values that Charter rights protect and promote, courts and administrative decision makers can assess whether restrictions on conduct or activities amount to serious, moderate, or minor limitations on rights. In the second place, seeing values and rights as tied closely together in this way alleviates the concern Côté and Brown JJ expressed in Trinity Western that Charter rights face some kind of threat from Charter values.

b) Values-advancing reasoning

A second way in which Charter values can inform an administrative decision maker’s limitations analysis is for the decision maker to ask, not which of the available options impairs rights less, but which realises Charter values more fully. To illustrate the point, think of those cases where an administrative decision maker is required by their empowering statute to work towards the achievement of a particular statutory objective.

If one course of action open to the decision maker would promote the achievement of that objective but would at the same time impose limits on a Charter right, the decision maker will have to determine if limiting the right is justified. In Trinity Western, there were only two options before the decision maker, and there was nothing to be gained by asking which of them impaired rights less: one of them obviously did. But knowing that a course of action that does not advance statutory objectives is less impairing of rights tells us nothing about the constitutionality of the course of action that does advance those objectives.

In Trinity Western, then, the majority and McLachlin CJ relied on interest balancing instead of minimal impairment as the mode of reasoning. In both sets of reasons, the judges considered how the LSBC’s decision balanced the adverse consequences of limiting TWU’s right to religious freedom on one hand, and the benefits of a government institution acting to protect the interests of members of the LGBTQ community on the other.

The difference in the two sets of reasons, however, highlights one of the shortcomings of interest balancing as a mode of reasoning. For the majority, it was easy enough to explain the balance between the limitation and its benefits, because they saw the rights limitation to be minor but understood the benefits achieved by limiting the right to be significant. The benefits outweighed the harms. To use the visual metaphor so often relied on in this area of jurisprudence, weighing the limitation’s significant benefits against its minor impact on rights tipped the scales in the favour of the limitation.

In contrast, the mechanics of McLachlin CJ’s logic are less obvious. Although she also saw great value in the LSBC protecting the LGBTQ community against discrimination in legal education, she found the limitation of TWU’s religious freedom rights to be extensive and serious rather than minimal. If it is intuitively right, or at least intelligible, for significant benefits to justify minor limitations on rights, interest balancing (and the balancing metaphor that accompanies it) does not offer any analytical clarity when significant benefits and serious rights limitations are weighed against each other. The scales, surely, are even. Once the Chief Justice had concluded that the extent of the rights limitation in Trinity Western was serious and that the benefits produced were significant, it is not clear how she nevertheless found the benefits to outweigh the costs.

If simply weighing up adverse consequences and salutary benefits does not produce an obvious or intelligible answer, judges and administrative decision makers must do more to explain their decisions. In R v KRJ, a majority of the Court (per Karakatsanis J) emphasised the importance of clear and accessible reasoning at the final stage of the Oakes analysis:

This final step allows courts to stand back to determine on a normative basis whether a rights infringement is justified in a free and democratic society. Although this examination entails difficult value judgments, it is preferable to make these judgments explicit, as doing so enhances the transparency and intelligibility of the ultimate decision. ⁷⁰

Although this passage extols the virtue of explanation, it does not actually describe a mode of reasoning that can supply those explanations. Happily, interest balancing is not the only mode of reasoning on which decision makers and judges can rely when inquiries into minimal impairment are of no help. What I call values-advancing reasoning encourages officials to consider which of the options available to them advances or promotes Charter values better, without having to engage in the kind of intuitive, subjective, and ad hoc weighing of costs and benefits that interest balancing sometimes involves. Indeed, Doré itself foreshadowed this analytical approach by enjoining decision makers to “ask how the Charter value at issue will best be protected in view of the statutory objectives.” ⁷¹

An example of this mode of reasoning can be found in the second of the Supreme Court’s 2018 decisions dealing with a rights-limiting administrative decision: Groia v Law Society of Upper Canada. ⁷² The facts of this matter are similar to those in Doré, in that they involve a law society’s imposition of disciplinary sanctions on a lawyer and a complaint by that lawyer that the sanctions unjustifiably limited his Charter right to freedom of expression.

In this case, the Law Society of Upper Canada’s disciplinary bodies found that in defending a criminal accused at trial, Mr. Groia’s spirited and vigorous oral arguments had crossed the line into incivility. Specifically, Mr. Groia had frequently accused the Ontario Securities Commission’s counsel, the prosecutor in the case, of abuse of process. ⁷³ The Law Society’s Appeal Panel confirmed the initial Hearing Panel’s finding that Mr. Groia had committed professional misconduct, suspended him from practice for one month and made a costs award against him of $200,000.

The Law Society is mandated by statute to advance the public interest, the cause of justice, and the rule of law, by regulating the practice of law in Ontario. ⁷⁴ Key to achieving these objectives is the setting and enforcing of standards of professional conduct, including a requirement of civility during criminal trials. In a majority judgment, Moldaver J explained that incivility can undermine the fairness and integrity of judicial proceedings in at least four ways. ⁷⁵

In the first place, “overly aggressive, sarcastic or demeaning courtroom language may lead triers of fact, be they judge or jury, to view the lawyer—and therefore the client’s case—unfavourably.” ⁷⁶ Second, incivility can distract lawyers, judges, and juries from the factual and legal merits of the case being litigated. It does not serve the interests of justice if a trial judge is preoccupied with policing lawyers’ conduct instead of focusing on the evidence and legal questions. Third, incivility and fractiousness may increase stress for complainants or witnesses, for whom a trial is already a stressful experience, compromising the probative value of their testimony or undermining their willingness to participate at all. Finally, incivility in the form of vitriol, sarcasm, and baseless allegations of impropriety may undermine public confidence in the administration of justice.

At the same time, lawyers’ obligations to conduct themselves with civility cannot be allowed to compromise their duty of resolute advocacy. Indeed, the Supreme Court has stated that a lawyer’s commitment to their client’s cause is a principle of fundamental justice, ⁷⁷ and the Charter right to fair trial implies that criminal defence lawyers be free to advocate vigorously on behalf of their clients. ⁷⁸

In an adversarial criminal justice system, the defence lawyer’s commitment to their client’s cause and the duty of resolute advocacy take on particular importance. Vigorous and fearless advocacy ensures that a criminal accused can mount a full defence and challenge all aspects of the case against them. The implication of this demand for resolute advocacy is that the speech of lawyers in criminal trials must not lightly be limited. Defence counsel must not fear recrimination or discipline at the hands of the law society for doing what the criminal trial process and the Charter demand in the first place.

The balance the Law Society was required to strike here lay between Mr. Groia’s right—and indeed his duty—to speak freely in defence of his client, and the requirement of civility during criminal trials. Both of the goods in the balance, however—civility in trial proceedings and freedom of speech in legal representation—serve the same underlying Charter value: namely, that criminal trial procedures are as fair and truth-seeking as they can be. Criminal trials in which defence counsel are free to mount resolute and fearless cases on the behalf of their clients, and that are conducted with civility, will tend to allow for factual and legal issues to be properly ventilated and considered by judges and juries.

Where the question arises as to whether it is justifiable to limit a lawyer’s freedom of speech in order to promote the objective of civility, a decision maker will benefit from adopting a values-advancing mode of reasoning and asking which course of action will best promote the underlying value. Indeed, values-advancing reasoning explains the majority’s decision in this case even though there is some analytical work that an inquiry into the availability of less restrictive means can do as well. The Law Society’s Appeal Panel could have considered, for example, whether there was a less severe punishment that it could have imposed on Mr. Groia.

But considering whether there are less restrictive punishments is, of course, consequent on a finding that someone has committed a punishable offence in the first place. The primary question for the Appeal Panel, and the question to which values-advancing reasoning is most applicable, is whether sanctioning Mr. Groia’s conduct as uncivil or affirming it as an example of resolute advocacy would better promote the underlying constitutional value.

Following Doré, the majority framed the analysis as an exercise in balancing. ⁷⁹ In conducting the exercise, the majority paid close attention to the ‘core values’ of freedom of expression identified in Keegstra and RJR-MacDonald, particularly the search for truth. A finding of professional misconduct is more likely to strike a proportionate balance between rights and statutory objectives, the majority reasoned, where the impugned conduct “lies far from the core values of lawyers’ expressive freedom.” ⁸⁰ On the other hand, “[r]easonable criticism advances the interests of justice by holding other players accountable.” ⁸¹

In other words, when a lawyer’s conduct at trial is on the line of incivility but does not help in the search for truth, it does not advance the values that a lawyer’s right to expressive freedom at trial is meant to serve. In that case, a lawyer’s incivility may in fact undermine the search for truth for all the reasons that Moldaver J outlined. The balance between rights and statutory objectives here is to be found by considering whether allowing the lawyer’s conduct or sanctioning it will better promote the integrity of the criminal justice system by promoting the search for truth. This is values-advancing reasoning.

As it turned out, the Court decided in Mr. Groia’s favour and overturned the Appeal Panel’s finding as unreasonable. ⁸² Key to the Appeal Panel’s consideration of the case was the fact that Mr. Groia’s allegations of abuse of process had been based on a mistaken understanding of relevant points of law. However, as the Appeal Panel noted and the Supreme Court emphasised, Mr. Groia made these errors in good faith: his sincerely held views on these points of law simply turned out to be wrong.

The Supreme Court concluded that the Appeal Panel had not found a proportionate balance between the statutory objective of legal professionalism and right to free speech during trial. It found the Appeal Panel’s responses to Mr. Groia’s conduct at trial to be excessive: “Branding a lawyer as uncivil for nothing more than advancing good faith allegations of impropriety that stem from a sincerely held legal mistake is a highly excessive and unwarranted response.” ⁸³

On its own, this is not a particularly helpful comment. It merely asserts that the Appeal Panel’s decision was excessive, without explaining why. We can dig a little deeper into the Court’s reasoning, though, to see that the question it is really trying to answer is whether the Appeal Panel had properly considered whether, and how, its decision would promote the underlying value of a fair and truth-seeking criminal justice process.

In this respect, the Court pointed out that the Appeal Panel appeared not to consider the broader consequences of branding Mr. Groia’s conduct uncivil and punishing him for it:

Finding a lawyer guilty of professional misconduct on the basis of incivility for making an abuse of process argument that is based on a sincerely held but mistaken legal position discourages lawyers from raising these allegations, frustrating the duty of resolute advocacy and the client’s right to make full answer and defence. ⁸⁴

It is not difficult to see how the Court uses values-advancing reasoning here, even if it does not specifically identify that as the operative mode of reasoning. The Court’s logic is that the Appeal Panel’s decision undermines rather than advances the search for truth at trial by discouraging lawyers from speaking up when they believe, sincerely, that procedures are abusive or unfair. This is what leads the Court to conclude that the decision was unreasonable, not the mere fact that it was excessive.

Indeed, it is no longer sufficient for decision makers or courts of review to blandly assert that the limitation of a right is excessive or that it is not. The culture of justification that Vavilov affirms, and the obligation to explain difficult value judgments that KRJ describes, require decision makers and courts to provide an intelligible and transparent explanation for why a decision is excessive. Values-advancing reasoning allows officials to do precisely this.

5. It’s Not All About Oakes: Preferring Three-Headed Justification Over Rights Objectivism

I have sought to explain how courts and decision makers can follow the Doré approach and attend to the proportionality of rights-limiting administrative decisions without relying on less restrictive means analysis. Moreover, I have explained why it might be necessary in at least some situations—where minimal impairment is not a fruitful avenue of analysis—to rely on other modes of reasoning about justifiability. If all Doré does, as I suggest here, is set out alternative modes of reasoning about rights limitations in some circumstances, why has there been such strong resistance to following the Doré approach when Charter rights are at stake?

In this final part of the paper, I suggest that the unwillingness to embrace Doré stems from a failure to see how it fits into the bigger picture of justification. That failure, moreover, is a function of the critics’ faith in formalism. While Doré outlines one way for decision makers to demonstrate that rights limitations are justified, it is not and does not claim to be the only approach, or even the approach that will provide the most transparent and coherent justifications in all cases. On the contrary, Doré itself recognises that a full Oakes analysis, including a consideration of less restrictive alternatives, may be the best way in some cases to show that a rights limitation is justified. What matters to a culture of justification is that decisions limiting rights are shown to be justified, and that the justifications that decision makers offer are intelligible and coherent.

Indeed, this is precisely what section 1 of the Charter requires. As LeBel J pointed out in concurring reasons in Multani, several years before Doré, the Supreme Court “has never definitively concluded that the s. 1 justification analysis must be carried out mechanically or that all its steps are relevant to every situation.” ⁸⁵ Oakes has never been the only way to apply section 1.

Doré’s critics nevertheless prefer a form of rights objectivism over a culture of justification. They insist that decision makers faithfully follow the full Oakes formula every time Charter rights are implicated. They are firm in their conviction that minimal impairment analysis is the only way for decision makers to objectively show that Charter rights have been correctly protected. But if Oakes is not the only way to apply section 1, and in some cases not even the most intelligible way of demonstrating whether or not a limitation is justifiable, the critics’ approach undermines a culture of justification by denying decision makers the opportunity to explain their decisions in the clearest and most accessible way.

In what follows below I explain why Oakes, and the rights objectivism that underlies the lionisation of minimal impairment analysis, is not categorically better at justifying rights-limiting decisions. I point to two flaws in the critics’ position, one doctrinal and the other theoretical. Before describing these two mistakes in more detail, it is useful to summarise the critical response to Doré.

a) The discomfort with Doré

Much of the concern with Doré’s approach to rights analysis revolves around the idea that an administrative-law based reasonableness review—even a ‘robust’ one—will underpower rights review. This criticism has come from both academic commentators and judges, and is succinctly summarised by Lauwers JA in a majority judgment in the Ontario Court of Appeal:

I would be reluctant to apply a robust concept of “reasonableness” burdened by a standing obligation of judicial deference to a line decision maker’s discretionary decision. There is a real risk that a claimant’s Charter rights will not be understood and will not be given effect by the line decision maker. ⁸⁶

The general tenor of this concern is a mistrust of administrative decision makers’ competence to properly balance Charter rights and statutory obligations. And perhaps it is unrealistic to expect what Lauwers JA calls a ‘line decision maker’ to have an understanding of Charter rights, proportionality analysis, and human rights theory as sophisticated as his own.

But the fact that line decision makers are not human rights theorists does not preclude them from being able to reason through the justifiability of a rights limitation, nor from explaining their conclusions on those questions. As the UK Supreme Court said in Denbigh (speaking then as the Judicial Committee of the House of Lords), line decision makers will sometimes be called on to make decisions that affect human rights but they “cannot be expected to make such decisions with textbooks on human rights law at their elbows.” ⁸⁷

At the same time, the recipients of line decision makers’ decisions tend not to be human rights theorists either. The explanation for a decision must be tailored to the audience to which it is given. A decision written in dense legal language is no more intelligible to a layperson than a decision that is poorly explained. Decision makers must be sensitive to the context in which they render decisions, just as much as courts of review must be sensitive to the context in which decision makers operate. Sometimes, as Vavilov points out, a common-sense explanation of an administrative decision—including, presumably, a rights-limiting decision—is exactly what the context requires:

A reviewing court must bear in mind that the written reasons given by an administrative body must not be assessed against a standard of perfection. That the reasons given for a decision do “not include all the arguments, statutory provisions, jurisprudence or other details the reviewing judge would have preferred” is not on its own a basis to set the decision aside: Newfoundland Nurses [2011 SCC 62; [2011] 3 SCR 708], at para. 16. The review of an administrative decision can be divorced neither from the institutional context in which the decision was made nor from the history of the proceedings.

Administrative decision makers cannot always be expected to deploy the same array of legal techniques that might be expected of a lawyer or judge—nor will it always be necessary or even useful for them to do so. Instead, the concepts and language employed by administrative decision makers will often be highly specific to their fields of experience and expertise, and this may impact both the form and content of their reasons. These differences are not necessarily a sign of an unreasonable decision—indeed, they may be indicative of a decision maker’s strength within its particular and specialized domain. “Administrative justice” will not always look like “judicial justice”, and reviewing courts must remain acutely aware of that fact. ⁸⁸

The Supreme Court has been quite clear, on the one hand, that administrative justice need not always look like judicial justice. Doré’s critics have proposed on the other hand that when Charter rights are involved, administrative justice must look like judicial justice. Decisions that engage Charter rights, they contend, must be correct. Moreover, the critics continue, Oakes and the inquiry into minimal impairment provide the framework for making correct decisions, while the Doré approach guarantees only reasonable decisions. ⁸⁹ What Doré’s critics demand most loudly, perhaps, is that administrative decisions affecting Charter rights be judicially reviewed on the standard of correctness rather than on the standard of reasonableness.

Doré’s critics, then, take heart in Vavilov’s formalism. The new standard of review test adopted in Vavilov categorically requires that all constitutional questions be reviewed on the standard of correctness. Since the question of whether a Charter right has been justifiably limited appears very much to be a constitutional question, it seems that there is no room left even for Doré’s robust reasonableness review of Charter rights-limiting administrative decisions. ⁹⁰

This argument misses the point, I think. Even if, after Vavilov, courts are supposed to review Charter-affecting administrative decisions on the standard of correctness, that has absolutely no impact on what Doré instructs administrative decision makers to do. Doré provides administrative decision makers with a range of modes of reasoning by which to inquire into whether or not a limit on a Charter right is justifiable, and explain the outcome to the subjects of their decision. Whether that decision is ‘correct’ or ‘reasonable’ is completely irrelevant. All that matters to the subject of the decision is that it be justified, and all that matters to the decision maker is that they are able to explain their decision to the people it affects.

Sometimes the best and most accessible explanation is given by proceeding through all the stages of Oakes, sometimes by interest balancing, and sometimes by values-advancing reasoning. What courts do on review is neither here nor there to decision makers’ efforts to explain their decisions as well as they can. And indeed, the more clearly and transparently decision makers are able to explain their decisions, the less likely the subjects of those decisions are to seek judicial review.

Moreover, the line between correctness and reasonableness when considering whether a rights limitation has been justified by the decision maker is, after Vavilov and the emphasis it places on a culture of justification, something of a red herring. Vavilov’s account of the difference between correctness review and reasonableness review makes nonsense of the idea that the latter is stricter or more exacting than the former. All that separates the standards of review after Vavilov is whose reasoning demonstrates that the outcome is justified: the administrative decision maker’s or the court’s. The majority in Vavilov (Abella J concurring on this point) said,

What distinguishes reasonableness review from correctness review is that the court conducting a reasonableness review must focus on the decision the administrative decision maker actually made, including the justification offered for it, and not on the conclusion the court itself would have reached in the administrative decision maker’s place. ⁹¹

Correctness review, it follows, involves a court of review considering for itself which outcomes of a decision-making process can be justified. There is no higher standard of justification here: whether it is a court coming to its own conclusion (on the standard of correctness), or a court reviewing a decision maker’s explanation for their conclusion (on the standard of reasonableness), an acceptable decision in either case is one that can be justified in an intelligible and transparent way. And in that regard Doré and the jurisprudence that follows continue to be relevant because they set out the full range of modes of reasoning by which any official—an administrative decision maker or a court of review—can consider the justifiability of a Charter rights limitation.

Doré’s critics are right, of course, in suggesting that the Oakes formula offers one way for administrative decision makers to explain their decisions about rights limitations. But insisting that Oakes is the only way to do so, and thereby rejecting Doré’s approach, puts the critics in the odd position of having to explain what Doré is actually about. In a valiant and understandable attempt to make sense of Doré’s vexing conflation of rights and values, the critics propose that judicial review should be bifurcated: where only Charter values are engaged it may be sufficient for decision makers to follow Doré’s approach and forego the minimal impairment analysis. Where Charter rights are limited, however, decision makers must resort to the more scrupulous analysis involved in the full Oakes formula, including an inquiry into minimal impairment. ⁹²

Mark Mancini suggests something along these lines in calling for a “stricter reasonableness review” where administrative decisions engage Charter rights. ⁹³ Without explicitly calling for a full Oakes analysis, Mancini recommends that the Supreme Court “could impose explicit reasoning requirements” where Charter rights are in play, arguing that Charter rights could be understood as elements of the legal constraints that bind administrative decision makers under Vavilov’s conception of reasonableness. ⁹⁴ “Reasoning about constitutional rights in this way,” Mancini concludes, “would require more than what Doré currently prescribes.” ⁹⁵

It is certainly true, as I acknowledge above, that Doré leaves much to be desired. It does not offer much in the way of guidance to administrators as to how to reason through whether a limitation is justified in those cases where less restrictive means analysis is not helpful. To that extent I agree with Mancini that reasoning about rights requires more than what Doré offers, at least explicitly. Values-advancing reasoning and interest balancing, therefore, are modes of reasoning that supply Doré’s omission in this respect. In what follows below, I explain more precisely why Doré’s critics are wrong to think that Oakes is the only option, emphasising instead the value of a three-headed approach to justification.

b) The doctrinal mistake: the dangers of rights objectivism

The critics’ first mistake is doctrinal. Doré itself makes clear that the approach it introduces is only meant to be used in quite narrow circumstances—that is, in the context of adjudicated decisions. The unanimous Court was careful to emphasise this, as is clear from the opening few paragraphs of the judgment. Doré’s case, the Court noted, “raise[d] squarely the issue of how to protect Charter guarantees and the values they reflect in the context of adjudicated administrative decisions.” ⁹⁶

Confronting the question of whether the presence of a Charter issue required the adoption of the Oakes test instead of the more traditional administrative law test for reasonableness, the Court went on:

It seems to me to be possible to reconcile the two regimes in a way that protects the integrity of each. The way to do that is to recognize that an adjudicated administrative decision is not like a law which can, theoretically, be objectively justified by the state, making the traditional s. 1 analysis an awkward fit.

…

[W]hile a formulaic application of the Oakes test may not be workable in the context of an adjudicated decision, distilling its essence works the same justificatory muscles: balance and proportionality.… In assessing whether an adjudicated decision violates the Charter … we are engaged in balancing somewhat different but related considerations, namely, has the decision-maker disproportionately, and therefore unreasonably, limited a Charter right. ⁹⁷

The decisions in Doré, Groia, and Trinity Western highlight that the Court is alive to the distinctiveness of adjudicated administrative decision making. In each of those cases, the decision maker was faced with just two options from which to choose. Each decision maker had to decide whether to take the decision that would advance statutory objectives, even though doing so would limit rights, or whether to refuse to take that decision because the rights limitation was unjustifiable.

This situation is quite different to that in which a legislature finds itself. Once the government has identified an important goal towards which to work, the legislature is free to construct any number of policy frameworks, from the ground up, through which to pursue that objective. Where there is a multitude of options and the resulting legislative decision will affect a diverse range of people in numerous ways, it makes sense to ask which of those options impairs Charter rights the least.

Not all administrative decisions are adjudicative, though. On the contrary, some are legislative. The common law of procedural fairness has long recognised that administrative decisions range in nature from more adjudicative to more legislative, and treats them differently. Where an administrative decision is legislative—for example where a public utilities regulator considers whether to approve the rates a telephone operator has set, ⁹⁸ or where a municipality passes a zoning by-law that applies generally to a city ⁹⁹ —the common-law duty of procedural fairness may not even arise. ¹⁰⁰

On the other hand, when an administrative decision looks like a court’s decision, the procedures by which it is made should approximate those of a court. Consider this classic statement from the Supreme Court’s important 1999 decision in Baker v Canada:

The more the process provided for, the function of the tribunal, the nature of the decision-making body, and the determinations that must be made to reach a decision resemble judicial decision making, the more likely it is that procedural protections closer to the trial model will be required by the duty of fairness. ¹⁰¹

This all goes to show that the common law is already quite willing to treat adjudicative and legislative administrative decisions differently. The recognition of this difference underlies the unanimous Court’s approach in Doré. When an administrative decision maker is seized of a legislative decision and must choose between the same kind of policy options as a legislature when considering and debating a bill, they may well benefit from considering which of those options is least restrictive of Charter rights.

But where the decision is adjudicative, where the decision maker has only a limited number of options open to them and these affect only a small number of people in foreseeable ways, asking which of those options is less restrictive is just not very useful. The confinement of the Doré approach to adjudicated decisions is important because it recognises and leaves room for less restrictive means analysis in other, more legislative contexts, where that inquiry will actually help the decision maker to reason through the justifiability of the available options.

Doré’s critics, nonetheless, insist on a kind of rights objectivism that demands all rights limitations be objectively proven to be minimally impairing. Only if a decision maker can demonstrate that a decision is the least restrictive way to achieve statutory objectives, they argue, can it be concluded that the decision is justifiable. The critics admit of no other form of explanation or argument that a rights limitation might be justified. When objective proof of minimal impairment is unavailable, the critics add, no rights limitation can be said to be justifiable.

But think again about Trinity Western: the LSBC concluded that refusing to accredit TWU’s law school would achieve the statutory objectives of promoting the public interest, while approving it would not. There was only one way for the LSBC to further its statutory objective here, even though that was not the least restrictive option available to it. To ask about minimal impairment, when the minimally impairing option fails to further statutory objectives, is the wrong question to ask.

The correct question—and the one that the culture of justification requires us to ask—is whether the option that does further statutory objectives but also limits rights is justified. The risk that the critics run in insisting that decision makers follow only the objectively least restrictive course, is in closing off even the possibility of explaining why rights limitations in adjudicated decisions might be justified by the advances they make in achieving statutory objectives. Indeed, that undermines the entire purpose of section 1 to the extent that it prevents administrative decision makers from fulfilling statutory objectives by taking decisions they would otherwise be able to explain as justified limitations on rights.

It is important to realise, then, that Doré does not close the door on less restrictive means analysis or on the rest of the Oakes analysis. Quite the opposite, Doré indicates that Oakes still has its place in Charter-engaging administrative decision making. The Doré approach—which as I have explained here can take the form of either interest balancing or values-advancing reasoning—stands alongside the full Oakes analysis as one mode of reasoning about the justifiability of rights limitations. More to the point, the Doré approach is particularly useful in precisely those situations where less restrictive means analysis does not offer intelligible answers. This is an important doctrinal distinction that, once appreciated, shields Doré from a great deal of the abuse it has taken.

c) The theoretical mistake: how proportionality in the strict sense works behind the scenes

Rights objectivism’s second mistake is theoretical, and it picks up where the doctrinal mistake leaves off. Even in the context of a legislative administrative decision, the minimal impairment inquiry can answer the question of justifiability only under one condition: namely, when all of the options open to the decision maker are equally effective in achieving the relevant statutory objectives.

Put simply, the course of action that impairs rights least, but which achieves the statutory objective as effectively as any other alternative, is plainly preferable. It is not justifiable to limit rights more than necessary to achieve a statutory objective, so when one option is more restrictive but no more effective, it cannot be justifiable to select it.

A recent Ontario Court of Appeal decision about changes to election law is illustrative. ¹⁰² The issue in Working Families was whether the limitation imposed on freedom of expression by extending the period of restricted political spending before a provincial election from six months to twelve months was justified. The legislative history was important here, because the Ontario government had already amended the legislation once before. In those earlier amendments, however, the government had left the restricted period at six months and its own experts had concluded “that a 6-month period of pre-writ regulation was reasonable.” ¹⁰³

The evidence suggested, then, that a six-month restricted period was just as effective in achieving the goals of fairer and more equitable elections as a twelve-month period. But as Morgan J noted, the government had nonetheless selected the more restrictive of these equally effective options. That, by itself, was enough for him to conclude that the limitation was not justified under section 1:

It is self-evident that if a six-month impairment on free expression accomplishes the desired objective, a twelve-month impairment cannot be the least drastic means. It does not matter that both a 12-month and a 6-month restricted period are “reasonable” in the view of the experts. A 12-month impairment is twice as long, and twice as restrictive, as a 6-month impairment, and so by definition is not minimal. There is nothing else to factor into the analysis. ¹⁰⁴

If a less restrictive alternative is also less effective in achieving the statutory objective, however, determining that it is less restrictive is not the end of the story. It must now be asked, with respect to both the less restrictive but less effective alternative and the more restrictive but more effective option, whether the extent of the limitation is justified by the benefits that it produces. The balance between the adverse and salutary consequences of each option is different in each case, and a consideration of the proportionality of that balance is required in both cases. Lebel J pointed this out in Hutterian Brethren:

The stated objective is not an absolute and should not be treated as a given. Moreover, alternative solutions should not be evaluated on a standard of maximal consistency with the stated objective. An alternative measure might be legitimate even if the objective could no longer be obtained in its complete integrity. At this stage of the proportionality analysis, the overall objective of the s. 1 analysis remains constant: to preserve constitutional rights, by looking for a solution that will reach a better balance, even if it demands a more restricted understanding of the scope and efficacy of the objectives of the measure. ¹⁰⁵

What the critics miss here is that balancing—either interest balancing or values-advancing reasoning—turns out to do a great deal of the analytical lifting even in those cases where an inquiry into less restrictive alternatives is analytically helpful. It is true that minimal impairment analysis provides a very good way to explain rights-limiting decisions in some limited circumstances, but it is not useful, by itself, in most other situations. In the situations where minimal impairment analysis cannot provide an objectively determinative answer to the question of whether a limitation is justified—as it did in Working Families—a decision maker must resort to persuasion, argument, and reasons: exactly what Vavilov’s culture of justification requires.

Of course, it is difficult to make an argument that a course of action which limits a right more than others, but realises a valuable goal more fully than others, strikes the best available balance between rights and broader social goals. For that matter, it is equally difficult to argue that a less restrictive but less effective course of action is more compliant with the Charter. But these are the difficult arguments that Vavilov’s culture of justification requires decision makers to make. As the Court said in KRJ, it is important that these difficult value judgments about the constitutionality of rights limitations be made transparent and accessible. ¹⁰⁶

Seen in the light of Vavilov, the Doré approach (interest balancing or values-advancing reasoning) and the Oakes formula (including minimal impairment analysis) are just different ways to make these value judgments explicit. The Court’s approach in Vavilov embraces the importance of persuasion and argument. What matters is not that administrative decisions comport with some idea of objective moral truth, but that the people affected by an administrative decision understand why an official made the decision they did. ¹⁰⁷

6. Conclusion

According to Greek mythology, Hercules was told by the Oracle at Delphi to serve the Mycenaean king Eurystheus, who in turn assigned Hercules twelve labours. Hercules was a reluctant hero, and would not have chosen to complete the labours were he not compelled to. Of these twelve labours, eleven involve slaying a beast (the Lernaean Hydra and the Nemean Lion, for example) or capturing or stealing something of value for Eurystheus (including the Cretan Bull, the golden apples of Hesperides, and the Mares of Diomedes).

The capture of Cerberus is different from all the other labours, however, because Hercules did not kill or steal him. Rather, after showing him to Eurystheus, Hercules simply returned Cerberus to the gates of Hades unharmed. ¹⁰⁸ The Ancient Greek world depended on balance between the living and the dead, and to maintain that balance, the Underworld needed Cerberus to guard its gates. After Hercules and Eurystheus were finished with him, Cerberus went back to work keeping that balance between the living and the dead, with all three heads intact.

My suggestion in this paper is that we understand administrative decision makers as playing a similar role to Cerberus, keeping balance between Charter rights and the compelling objectives the legislatures might want the administrative branch to pursue. Just as Cerberus uses his three heads, with three sets of eyes and teeth, to maintain balance between the Underworld and the living, administrative decision makers can deploy three different modes of reasoning to determine whether it is justifiable to limit Charter rights in pursuit of statutory objectives.

Administrative decision makers are inevitably going to have to make decisions that engage Charter rights, and the culture of justification that Vavilov entrenches in Canadian public law demands that they justify any decisions that impose limits on Charter rights. It seems plain that affording decision makers as much opportunity as possible to provide justifications for their decisions is consistent with a culture of justification. Each of the modes of reasoning I describe in this paper offers decision makers a way to explain their decisions, and each is better suited to providing compelling and intelligible explanations in specific circumstances.

I worry that Doré is under threat, and that the balancing approaches to reasoning about limitations it espouses will be rejected in favour of a requirement that administrative decision makers and courts of review engage in a minimal impairment analysis every time rights are implicated. To me, that would only have the effect of eliminating two modes of reasoning by which administrative decision makers can think through the problems they face.

Just as cutting off two of Cerberus’s heads would leave him a less effective guard dog at the gates of Hades—a less effective keeper of balance between life and death—excising the Doré approach from Canadian public law would reduce the analytical resources available to administrative decision makers and undermine the prospects of a culture of justification. My plea, then, is that we recognise the value that interest balancing and values-advancing reasoning have as modes of reasoning, and retain them alongside minimal impairment analysis as means by which decision makers and courts can, in appropriate circumstances, explain why they find a decision to impose justifiable or unjustifiable limits on Charter rights.

A culture of justification surely demands as much justification as possible.