Corps de l’article

introduction

LEGISLATURES routinely delegate broad powers to the executive to make regulations and other rules that affect our lives.[1] What role should the courts play in reviewing these rules? Traditionally, Canadian courts have limited themselves to determining whether rules were formally authorized by legislation. In Katz Group Canada Inc. v. Ontario (Health and Long-Term Care), the Supreme Court of Canada said it was not for the courts to inquire into the “political, economic, social or partisan considerations” that might have influenced rules’ design, nor to consider whether rules “will actually succeed at achieving” their underlying objectives.[2] On this view, arguments by affected parties about the expected costs and benefits of new rules are to be weighed exclusively by administrators. By contrast, judicial review of the substantive reasonableness of adjudicative decisions by administrators has long been viewed as uncontroversial. As the Supreme Court explained in Canada (Minister of Citizenship and Immigration) v. Vavilov, “rationality and fairness” demand that those who exercise “delegated public power,”[3] among other things, “meaningfully account for the central issues and concerns raised by [affected] parties.”[4] For example, an adjudicator may need to respond to concerns that a decision will fail to achieve its intended purpose or impose burdens disproportionate to its benefits.[5]

Prior to Vavilov, the Supreme Court issued a series of decisions suggesting that adjudication and rulemaking are both subject to judicial review for reasonableness.[6] Applying this standard, as defined in Vavilov, would mean assessing a rule’s compatibility not only with its enabling statute but also with the factual record before the rulemaker (including submissions by affected stakeholders).[7] Commentators and some appellate courts have argued that reviewing all administrative decisions under the Vavilov framework would make for more coherent doctrine.[8] Others have resisted this push, including the Court of Appeal of Alberta; they similarly point to doctrinal grounds for treating rulemaking differently from adjudication.[9]

Rather than focusing on doctrine, this article examines a perhaps more compelling reason to resist stringent review of rulemaking: its potential consequences for stakeholders. On this rubric, the idea of subjecting rulemaking to reasonableness review might seem unobjectionable at first glance. If rulemakers need to justify their decisions in light of available evidence and submissions by affected stakeholders, they are less free to engage in a number of problematic behaviours. These include shirking (promulgating rules without doing much work to understand their likely effects),[10] pandering (promulgating rules that are popular, but that the public would not prefer if they had the time and resources to subject these rules to closer examination),[11] as well as deliberate catering to vested interests.[12] In Canada, where rulemakers generally have been under little to no obligation to consider stakeholder interests before promulgating new rules, the risks of these behaviours are especially salient.[13]

But reasonableness review can also serve as a vehicle through which vested interests gain disproportionate influence over the rulemaking process, especially when rulemakers are tasked with protecting a diffuse group with little time or incentive to participate in the rulemaking process (e.g., consumers or workers) by imposing burdens on a smaller group of actors with considerable incentive and resources to participate in this process (e.g., a regulated industry).[14] As the United States’ experience with reasonableness review of rulemaking illustrates, regulated industries are well placed to present sophisticated projections of the costs of new rules and to point out apparent flaws or sources of uncertainty in rulemakers’ estimates of the potential benefits of these rules.[15] Responding to these kinds of submissions with countervailing analysis is costly for the rulemaker—perhaps prohibitively so.[16] As such, to the extent reasonableness review is taken by courts to require rulemakers to respond convincingly to these kinds of submissions before adopting new rules, it may serve to embed a status quo bias in regulation that operates to the detriment of the public.[17] Rather than being proactive and addressing potential harms before they occur, a rulemaker operating under such a regime may tend to intervene only after harm occurs and hence is easy to measure and prove.[18]

The adoption of reasonableness review need not lead to these consequences, however, provided it is applied in a way that recognizes a vital difference between adjudication and rulemaking. Adjudication is concerned largely with past facts, which once found allow for the application of substantive rules to a standing controversy.[19] By contrast, rulemaking is concerned largely with future consequences, which are considerably more uncertain.[20] This uncertainty makes it difficult, perhaps at times impossible, for rulemakers to prove that their view of the likely effects of new rules is more compelling than views advanced by stakeholders.[21] Courts should not expect the impossible. When the record is capable of supporting multiple views about the effects of new rules, courts should give rulemakers space to make value judgments about how to proceed in the face of this uncertainty. This is justified not on account of rulemakers’ expertise—they are no more qualified to make value judgments than others—but rather their accountability to more democratic branches of government.[22]

Deference to these value judgments need not last forever.[23] Courts engaged in reasonableness review could look to whether a rulemaker has committed to review the effects of its actions after some period and hold the rulemaker to this commitment. In this way, rather than serving to ossify regulation by requiring upfront, comprehensive consideration of every potential consequence of new rules, reasonableness review could encourage rulemakers to continually learn and adapt to new evidence of the effects of their rules. If reasonableness review is allowed to take this form, concerns about the consequences of its adoption could be significantly abated.

Part I of this article reviews the doctrinal debate over reasonableness review. Part II sheds light on the potential consequences of adopting reasonableness review by examining the role and relative power of different interest groups in the rulemaking process. Part III expands on the alternative framework for reasonableness review described earlier. Part IV considers how this framework could shed light on two related issues that courts seem likely to revisit if they adopt reasonableness review of rulemaking: whether guidance also ought to be subject to substantive review for reasonableness,[24] and whether rulemaking and guidance alike ought to be subject to review for procedural fairness.[25]

I. THE DOCTRINAL DEBATE

Rules have long been reviewable for vires. As framed by Justice Abella in Katz Group, this means examining whether a connection can be drawn between the rule and its enabling statute. The rule must be authorized under the terms of the statute and must not be “irrelevant,” “extraneous,” or “completely unrelated” to the objectives behind that authorization.[26] Whether a rule meets this standard is nominally a question of law on which no deference is owed to the rulemaker, but courts are instructed to apply a “broad and purposive approach”[27] to interpreting enabling legislation so that rules can be upheld “where possible.”[28] The standard does not allow courts to probe into the “political, economic, social or partisan considerations” that might have influenced a rule’s design or consider whether the rule “will actually succeed at achieving the statutory objectives.”[29] Nor may courts invalidate rules merely on the basis of perceived shortcomings in the reasoning process that led to their adoption, including a perceived failure to respond to arguments advanced by affected parties (whether before or after their adoption).[30] Like legislators, rulemakers are free to ignore arguments from affected parties, take partisan political considerations into account, and rely on evidence others might consider flawed.[31] As long as the rules they produce are authorized by and at least notionally promote the objectives of their underlying statute, the court is to uphold them.

The vires test was thought to exist outside the standard of review framework for substantive review of administrative decisions, but a trilogy of Supreme Court decisions over the 2010s highlighted how these regimes can run together.[32] In Catalyst Paper Corp. v. North Cowichan (District), the Court applied the standard of review framework in respect of a municipal by-law.[33] In Green v. Law Society of Manitoba, a majority of the Court applied this framework to rules promulgated by a law society.[34] And in West Fraser Mills Ltd. v. British Columbia, a majority of the Court applied it to rules made by a workers’ compensation board.[35] In each case, a standard of reasonableness was applied and the rules at issue were upheld.[36] But this approach has not been unanimously embraced by the Court: in West Fraser Mills, various minority opinions argued that rules ought instead to be reviewed on a standard of correctness (given that, under a traditional vires approach, the court does not defer to rulemakers’ interpretations of their enabling legislation).[37]

And, indeed, one could be forgiven for assuming that this apparent shift from vires to reasonableness entails a more deferential approach to the review of rulemaking.[38] But Chief Justice McLachlin’s majority opinion in Catalyst Paper illustrates how, in this context, reasonableness review counterintuitively serves to elevate the standard applicable to the judicial review of rules. Reasonableness review takes the traditional vires test and reads in an additional limit to the scope of authority deemed to have been given to the rulemaker: not only must rules be formally authorized by enabling legislation and connect to the objectives underlying that authorization, but they must also be a reasonable exercise of the authority delegated to the rulemaker.[39]

This limit, Chief Justice McLachlin says, derives from “the fundamental assumption ... that a legislature does not intend the power it delegates to be exercised unreasonably.”[40] If rules were “partial and unequal in their operation as between different classes; if they were manifestly unjust; if they disclosed bad faith; if they involved such oppressive or gratuitous interference with the rights of those subject to them as could find no justification,” they should be struck down unless legislation expressly contemplates such action.[41] This is because absent such language, we cannot presume the legislature “intended to give authority to make such rules.”[42]

Is reasonableness review now available for all rules, whatever their source? Vavilov seems to suggest so, casting its new standard of review framework as a “unified approach” to judicial review of “all types of administrative decision-making.”[43] If so, then the new, more stringent reasonableness standard it outlines could have significant implications for rulemaking. Vavilov contemplates—in addition to checks for legal invalidity, bad faith, and unjustifiable consequences—a close review of the submissions and evidence before the rulemaker, as well as the rulemaker’s asserted policy basis for its rules, to see whether the rules connect to these “factual constraints.”[44] If the rule “is in some respect untenable” in light of these constraints—for example, because it reflects a failure by the rulemaker to “meaningfully account for the central issues and concerns raised by [affected] parties” or the drawing of “conclusions [that] were not based on the evidence that was actually before [the rulemaker]”—then it is unreasonable and therefore invalid (i.e., ultra vires).[45]

Yet doubts about whether it makes doctrinal sense to regard the Vavilov framework as extending to the judicial review of rulemaking linger. John Evans cites the potential for “confusion” on the part of courts engaging in a Vavilov-style reasonableness review of rulemaking.[46] Courts are accustomed to applying a reasonableness standard to tribunal decisions, where the standard “connot[es] judicial restraint.”[47] Accordingly, they may misread the “reasonableness” label as a signal to defer to rulemakers on the same matters on which tribunals have historically been owed deference, like their interpretations of the boundaries of their authority.[48] In Evans’s view, the lack of procedural guardrails around the rulemaking process militates against deferring to any legal interpretations developed within this process.[49] Evans also suggests that “the central importance that Vavilov attaches to the reasons given by a decision-maker” may give rise to further problems, given that rules often are not accompanied by formal reasons.[50] In his view, these and other potential sources of confusion make the Vavilov framework a “poor fit” for the judicial review of rulemaking, such that its scope should be confined to the adjudicative context.[51]

As John Mark Keyes observes, however, adjudicative and rulemaking functions are not watertight compartments.[52] Take, for example, courts’ rationale for deferring to tribunals on the interpretation of their home statutes. This deference flows not from the procedures tribunals follow, but out of recognition that “[t]he application of legislation cannot be divorced from its interpretation.”[53] Rather than being asked merely to apply legal rules, tribunals often are asked to adjudicate disputes in accordance with broad legislative objectives.[54] In this context, legal and policy questions blend together, such that rather than being left with a single, legally correct decision, a tribunal often must choose from among a range of valid, reasonable choices.[55] This is the same kind of choice that faces rulemakers. In both contexts, deference gives administrative actors freedom to set policy within the boundaries defined by legislation. This is perhaps why Katz Group tells courts to take a “broad and purposive approach” to interpreting rulemakers’ enabling legislation[56] such that rules can be reconciled with this legislation “where possible.”[57] This approach may not represent deference in quite so many words, but it seems pretty close.[58]

Keyes also notes, as highlighted above, that Vavilov contemplates scenarios in which decisions are not accompanied by formal reasons.[59] Here, the rationale for an administrative decision often can be deduced from “the record and the context”[60] without having to resort to review based solely on a decision’s outcome.[61] In rulemaking, total evidentiary vacuums likely would be rare: rulemakers often publish evidence supporting their final work product and consider submissions from affected parties;[62] in some fields, these actions are even required by legislation.[63] As such, Keyes sees little reason to leave rulemaking outside the Vavilov framework—if anything, in his view, retaining distinct approaches to the review of rules and tribunal decisions may create confusion rather than avoid it.[64]

Nonetheless, this doctrinal dispute has spilled over into Canadian appellate courts. In Auer v. Auer, the Court of Appeal of Alberta refused to apply Vavilov to the review of a rule because “judicial review for reasonableness relates to administrative acts, not law-making,”[65] and rulemaking is an outgrowth of the lawmaking process.[66] Any attempt to review rules for reasonableness would amount to courts’ “enter[ing] the legislative field by weighing in on ... economics, policy, motives for passing the regulation, the wisdom of or likely efficacy of the regulation, or the regulation’s impact on particular individuals.”[67] Other courts have taken the contrary position, however, and at time of writing the matter was under consideration at the Supreme Court of Canada.[68]

There are at least three reasons to expect the Court will ultimately weigh in against Auer.[69] First, casting rulemaking as merely an outgrowth of the legislative process understates its importance.[70] Rules don’t just fill gaps left by statutes. Rules can eclipse their enabling legislation in terms of their importance to regulated actors, often by design. For example, in a principles-based regulatory regime, legislation may limit itself largely to setting broad social objectives, and leave rulemakers to create and continually update rules designed to achieve these objectives.[71] Within such a regime, one can expect not only that rules will dwarf statutes, but also that the need to keep rules up-to-date means the rulemaking process will continue long after a statute is enacted.[72]

Second, given that rules, like tribunal decisions, are administrative acts that are supposed to promote social objectives set by legislatures, it would seem to make sense to review both categories of decision on the same footing. The Supreme Court has emphasized the ongoing “policy development role” tribunals play when making decisions—decisions that, like rules, may be rendered long after their enabling legislation is enacted.[73] And while the Court’s 2018 decision in Mikisew Cree First Nation v. Canada (Governor General in Council) held that the legislative process is not subject to judicial review, it made clear that it does not regard rulemaking as part of this process.[74] Rather, the Court grouped rulemaking and tribunal decisions together as decisions taken “pursuant to statutory powers” (on Justice Karakatsanis’s formulation),[75] or following royal assent to legislation (on Justice Brown’s formulation),[76] that lie outside the legislative process and accordingly are subject to judicial review.

Third, it does not seem inevitable that reviewing rules for reasonableness will “descend into” the kind of policy analysis warned of in Auer.[77] Reasonableness review does not entail second-guessing the merits of rules. Rather, it involves an exercise for which courts are much better equipped: the assessment of whether an administrative decision reflects the application of expertise and care, including meaningful engagement with concerns raised by affected parties.[78] In undertaking this work, courts can look to analyses published during the rulemaking process or (alternatively) rationales offered by government counsel on judicial review. To the extent rulemakers appear to exercise authority without any discernible rationale, or without accounting for submissions by affected stakeholders, they fail to live up to the expectations implied by the legislature’s decision to delegate authority to them.[79] Reasonableness review would allow courts to intervene when this happens.

So why keep resisting the application of Vavilov to the judicial review of rulemaking? Perhaps the reason has less to do with doctrinal tidiness than with potential consequences. Auer itself gestures at these consequences when it says “[t]hose seduced” by the idea of applying Vavilov to rules “will perhaps view increased constraints on government’s ability to create law and an expansive opportunity to challenge government policies in court as laudable.”[80] And in the United States, where courts abandoned vires review of rules in favour of reasonableness review decades ago,[81] a long line of administrative law scholarship lends credibility to these concerns. It argues that reasonableness review has been mobilized in the United States “to frustrate [rulemakers’] efforts to curb market exploitation, protect workers, and press for a fairer distribution of resources.”[82]

But how might reasonableness review give rise to these consequences? To answer this question, we need to look at how more stringent standards of judicial review might bend the rulemaking process in ways that benefit some interest groups over others.

II. interest groups and judicial review

Regulation invariably creates winners and losers: the benefits it creates for some stakeholders come at least in part from the costs it imposes on others.[83] Typically, the beneficiaries of regulation are some large, relatively vulnerable group viewed as a proxy for the public (e.g., consumers or workers), and the cost-bearers are some smaller group (e.g., a regulated industry).[84] The cost-bearers often are relatively well-resourced and informed players that, absent regulation, would have greater scope to use their advantages to exploit the beneficiaries.[85] But this is not always the case—in the immigration, prisons, and national security contexts, for example, beneficiaries and cost-bearers alike may be cast as vulnerable.[86]

Sometimes, legislation does most of the work of allocating costs and benefits among these winners and losers, with rulemakers left to fill in relatively minor details.[87] The discussion that follows is concerned largely with a different kind of context: where legislation leaves rulemakers with a broad scope to make rules in the public interest or in service of a set of high-level social objectives. Here, the stakes for beneficiaries and cost-bearers, as well as the potential costs of regulatory failure, seem prone to be more significant and hence more worthy of discussion.

Regulation might fail to benefit the public for any number of reasons. Rulemakers might shirk their responsibilities by not doing enough to assess whether their work product is likely to achieve its underlying objectives, saving themselves time and effort at the expense of everyone else.[88] Or they might pander to beneficiaries by creating rules that seem at first glance to promote their interests, but that they would not have chosen if they were fully informed of these rules’ likely effects relative to alternatives.[89] This might allow rulemakers to maximize their political support, again at the expense of everyone else. What is more, cost-bearers might use political donations or lobbying to lead rulemakers to place greater weight on their interests relative to those of the public.[90]

Reasonableness review can serve to mitigate each of these risks. If rulemakers know their rules will need to be justified with reference to their enabling legislation and the evidence placed before them (e.g., submissions from affected parties), that rulemaker has less scope to shirk, pander, or cater to vested interests when developing new rules.[91] This would seem to provide a rationale for the United States Congress’s decision, via the Administrative Procedure Act (APA), to empower US federal courts to invalidate rules where the record reveals they are “arbitrary” and “capricious.”[92] Rules fail this test “if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.”[93] Failures by rulemakers to “provid[e] an adequate rebuttal” to “vital” comments made by a stakeholder can cause rules to fail this test.[94] And as mentioned earlier, substantive reasonableness as outlined in Vavilov similarly hinges on whether an administrative decision-maker has “meaningfully account[ed] for the central issues and concerns raised by [affected] parties” and rooted its decision in “the evidence that was actually before” it.[95]

Seeking judicial review of government decisions is expensive, however, and regulatory beneficiaries and cost-bearers are not likely to have the same incentive to choose this avenue. Assuming beneficiaries are a larger group, it will be more difficult for them to organize and coordinate their efforts than is the case for cost-bearers.[96] And assuming the costs of ineffective regulation are more widely dispersed among this group, they will see less value in undertaking these efforts.[97] This prediction seems to be borne out in the US experience with judicial review, where evidence suggests industry groups are responsible for the bulk of applications for judicial review of economic and social regulation.[98] The same goes for the rulemaking process that precedes judicial review: typically, industry will be responsible for the bulk of the submissions placed before a US rulemaker.[99]

When it comes to the risks of shirking or pandering, this imbalance may not be a problem for beneficiaries. Cost-bearers’ efforts to challenge ineffective rules that benefit the government at the expense of everyone else will, by definition, yield positive results for beneficiaries as well. Of course, cost-bearers’ incentive to push against government action does not depend on its effects on beneficiaries: they can save costs by challenging effective and ineffective rules alike.[100] However, any negative implications for beneficiaries may be muted assuming government has the resources to push back against ill-founded challenges to its work product. In contexts where cost-bearers are not especially well-resourced (e.g., when it comes to rules on immigration or national security), this might be a reasonable simplifying assumption to make.

But the US experience with judicial review suggests that this assumption is anything but safe to make when cost-bearers have considerable resources at their disposal. Regulated industries can use these resources to exert considerable gravitational force over regulatory action—not despite the presence of judicial review, but because of it.[101] Industry can use the rulemaking process to flood the record with a slew of seemingly “vital” comments that, whether by emphasizing a proposed rule’s possible costs or highlighting uncertainties about its possible benefits, threaten a proposed rule’s likelihood of surviving review.[102] Rulemakers are especially vulnerable to this tactic when their enabling legislation leaves them with open-ended discretion in crafting rules—after all, the broader the discretion, the broader the universe of factors stakeholders could argue ought to figure into the exercise of that discretion.[103]

The “thousands of pages of unstructured, highly technical information”[104] included in these comments are not only taxing for rulemakers to review and respond to, they may not even be all that useful in evaluating rule proposals.[105] Industry comments may be based on flawed or even biased assumptions—one review of industry comments on environmental rulemaking found that industry tended to overestimate the costs of new rules.[106] While rulemakers may realize that these analyses ought to be viewed with skepticism, “they cannot reliably anticipate which [stakeholder] comments a reviewing court might someday find vital.”[107] As such, they “have little choice but to respond, often in punitive length and detail, to all the substantive comments they receive.”[108]

It stands to reason that rulemakers will not have the resources to undertake this work for every rule proposal they view as likely to benefit the public. To the extent this is the case, they may abandon promising rule proposals merely because they lack the staff and budget to conclusively rebut arguments submitted by industry.[109] Or they may place these proposals on the backburner until a more concrete case for regulation emerges—for example, a financial regulator might wait until after a financial crisis to push forward with policy proposals previously resisted by industry.[110] Delays and retreats in the face of credible threats of judicial review have been said to explain “a great deal about why so many [US] federal laws designed to protect consumers, workers, or the environment have been so ineffective.”[111]

For those dissatisfied with the general lack of procedural and substantive guardrails around rulemaking in Canada, this scenario would seem disheartening. Under reasonableness review, vested interests lose opportunities to collude with government to create new rules that benefit them,[112] but they gain opportunities to leverage judicial scrutiny of rulemaking to frustrate public interest-oriented rule proposals.[113] Might the US experience with reasonableness review be inapplicable in Canada? While certain legal and cultural differences might lead us to think reasonableness review would take a different direction in Canada, there is reason to doubt the significance of these factors.

To start with legal differences, the APA generally requires US rulemakers to undertake a notice-and-comment process before implementing new rules.[114] This involves giving public notice of a proposed rulemaking and then receiving and considering comments from the public before deciding whether to move forward with the proposal.[115] By contrast, Canadian rulemakers generally are under no legislative duty to consult the public before creating new rules.[116] However, if rules are promulgated without consulting with stakeholders, these stakeholders are free to make submissions requesting that the relevant rulemaker reconsider these rules.[117] And as the Federal Court of Appeal has pointed out, a failure to act on these submissions could form a basis for seeking judicial review on the Vavilov reasonableness standard.[118] So under Vavilov, Canadian and US rulemakers may differ only in terms of when they must consider and respond to stakeholder submissions.

In addition to public consultation, the APA requires a form of reason-giving: rulemakers must publish “a concise general statement of [their rules’] basis and purpose.”[119] Vavilov, by contrast, contemplates that some administrative decisions will be rendered without reasons or a record.[120] But again, this may only affect when rulemakers must offer justifications for their policy choices in response to challenges from industry. Suppose a court is engaged in reasonableness review of a rule without the benefit of reasons or a comprehensive record but has submissions from the challenging stakeholder explaining why it views the rule as unreasonable. Presumably, the court will expect the rulemaker’s counsel to offer a justification for the rule—it is difficult to see how it would be acceptable, post-Vavilov, for counsel to simply say that the rule was authorized by legislation and that rulemakers are under no duty to consider submissions that their rules are ineffective, produce costs disproportionate to their benefit, or are otherwise unreasonable. And in any event, the litigation process is likely to reveal at least some of the information before the rulemaker.[121]

To move briefly to possible cultural differences, the conjecture that legal culture[122] is more litigious in the United States than in Canada has long been a topic of discussion.[123] Perhaps this means that even if reasonableness review becomes available, Canadian industry players will prefer cooperation with the state over confrontation. Indeed, the principles-based approach taken to securities and other areas of financial regulation is predicated on the idea that regulators and industry will cooperate to achieve outcomes set by legislation.[124] Such norms need not be a function of altruism on the part of industry—to the extent they believe antagonizing their regulator will impair their interests, they would have reason not to be viewed as stepping too far to challenge regulatory action.[125] A similar dynamic is said to exist in the United Kingdom, where courts’ approach to judicial review of rulemaking is similar to the Katz Group approach.[126]

This leads naturally to what is probably the most important problem with arguments rooted in differences in legal culture: perhaps legal culture is itself a function of legal rules (at least in part), such that when the latter changes, the former will change to some extent as well.[127] After all, if industry believes courts are likely to defer to rulemakers, there is little to be gained from challenging their decisions. And to the extent any retaliation on the part of rulemakers is itself likely to be immune to challenge, that gives industry even more reason to pursue a cooperative approach. Once these conditions fall away, industry presumably will see more value in using the courts to challenge regulatory action.

Does this mean we should, as the Court of Appeal of Alberta did in Auer, reject the idea of subjecting rules to reasonableness review? Is this standard, despite seeming to work well in the adjudicative context, destined to do more harm than good in the rulemaking context? Or can reasonableness review be designed in a way that mitigates risks that rulemakers will engage in shirking, pandering, or catering to vested interests, while also mitigating the risk that regulated industries will use it to frustrate rules designed to serve the public? The next Part addresses this possibility, which requires that we revisit some larger questions about the relationship between courts and administrators.

III. reframing reasonableness review

The standards applied in Katz Group and Vavilov each emphasize a different response to the question of why courts should defer to rulemakers and other administrators. Katz Group emphasizes that courts should defer because the legislature chose to delegate authority to rulemakers, not courts.[128] As such, the court’s role on judicial review is to ensure the rulemaker has acted within the scope of the authority given to them by the legislature, but nothing more.[129] On this view, as expanded upon in Auer, inquiring into the substantive reasonableness of rulemaking would infringe on parliamentary sovereignty and the separation of powers between legislatures and courts.[130]

From the standpoint of Vavilov, this response dodges another question: why do legislatures delegate authority to rulemakers? On its analysis, if we exclude purely self-serving reasons, like avoiding blame for policy failures or giving political cover for unpopular rules, we are left with one obvious candidate: expertise.[131] Legislatures set out to define broad social objectives and then rely on rulemakers to develop and use expertise to figure out how to promote those objectives.[132] Even when rulemakers are not experts themselves, such as when a cabinet or minister acts in this capacity, they typically will have the benefit of informed analyses and advice from civil servants before taking action.[133] Accordingly, Vavilov points us toward the proposition that courts should defer to rulemakers only to the extent they draw upon expertise when designing rules.[134]

The trouble with emphasizing expertise as the justification for deference to rulemakers, however, is that it implies all problems raised by rulemaking can be resolved through the application of expertise. While rulemaking undoubtedly involves the application of expertise, it would be wrong to say this is all it involves—even more so than it would be to make the same claim about adjudication. This is not because adjudication involves law and rulemaking involves policy. As mentioned earlier, legal and policy questions pervade both of these functions.[135] Nor is it because one is a court-like process and hence amenable to judicial review and the other a legislative process that courts lack the expertise to review. Rulemaking and adjudication alike produce a record that allows courts to assess the substantive reasonableness of decisions.[136]

Rather, the key difference between these functions has to do with how much their respective decision-makers know about the factual constraints on their decisions at the time they make these decisions. Adjudicators typically deal in large part with facts that have already happened—the facts that led parties to seek adjudication in the first place.[137] While adjudicators might have to consider potential future consequences of their decisions, like the potential deterrent effect of a fine, this analysis draws heavily from the concrete facts that created a need for their intervention. For example, while a securities tribunal’s powers to make orders in the public interest (e.g., fines and bans on service as an officer or director of a public company) “are preventive in nature and prospective in orientation,”[138] their use hinges on the “past conduct” of the proposed subject of the order, which must be “so abusive as to warrant apprehension of future conduct detrimental to the integrity of the capital markets.”[139]

Rulemakers, by contrast, are concerned largely with the future. Past problems might have created a perceived need for new rules, but rulemakers are concerned with whether their rules will create social benefits by resolving these problems—and how likely it is that these benefits will exceed the associated costs. And unlike the adjudicator who deals with a static set of facts, the rulemaker operates in a dynamic environment: it creates rules, which in turn have effects on stakeholders, which may in turn create a perceived need for revisions to these rules.[140] As such, whereas adjudicators deal largely with facts, rulemakers deal largely with predictions—estimates of costs and benefits that may or may not turn out to be accurate, and hence should not be treated as gospel.[141] In the face of these uncertain estimates, rulemakers might want to take a precautionary approach: impose new rules before there is a clear case that their benefits will exceed their costs, and adjust their approach as they learn more about the effects of these rules.[142] And where rulemaking power is divided among multiple decision-makers—as is standard in city councils, but by no means unique to them—rulemakers may feel a need to make compromises based on differing views about the same uncertain situation.[143]

The principles that follow set out a pathway toward an approach to reasonableness review of rulemaking that gives rulemakers scope to make these kinds of decisions, acting under the guidance of their enabling legislation and the government of the day. These principles recognize the importance of expertise to administrative decision-making while also acknowledging the very real limits to this expertise—limits imposed by uncertainty about the future. In doing so, they meld the two justifications for deference emphasized in Katz Group and Vavilov: rulemakers apply expertise to resolve problems, but their link to more democratic branches of government also entitles them to make value judgments in response to uncertainty.[144] As such, the discussion that follows draws on some of the ideas sketched out in Katz Group, while also explaining these ideas’ continued relevance following Vavilov.

A. The Limits of Estimates

Neither rulemakers nor stakeholders have perfect foresight of the effects of new rules. No matter the sophistication of their respective analyses of these future effects, these analyses should be taken by courts for what they are: estimates and predictions, which inevitably intertwine objective measurement and more subjective value judgments.[145] The costs of harms caused by toxins or chemicals to the environment and human health are not readily quantifiable, and hence difficult to weigh.[146] The costs and benefits of new rules affecting financial activity may be easier to quantify, but the likelihood that affected actors will change their activities in response to these rules leads to more possibilities about what these costs and benefits might turn out to be.[147] This may make it impossible for a rulemaker, even if well-resourced and well-staffed, to demonstrate that its analysis of the costs and benefits of a new rule is better than that put forward by an affected stakeholder.[148]

Reasonableness review should not require rulemakers to achieve the impossible. Rather than force rulemakers to resolve all sources of uncertainty or disagreement about a new rule’s effects before acting, it should give rulemakers scope to test, fail, and learn from experience. What does this mean for a court reviewing stakeholders’ arguments that a rulemaker failed to adequately engage with their submissions? The court could start by asking whether the stakeholder’s submissions show that the rulemaker relied on premises that were wrong or implausible based on available evidence, or that the rulemaker reached conclusions unconnected with available information or any stated value judgments (e.g., a preference to err on the side of precaution by responding to potential risks before firm evidence of harm arises). If so, a failure to account for these submissions would reflect unreasonableness. For example, if the rulemaker were to assert without explanation that the benefits of a new rule are likely to outweigh its costs, while failing to engage with contrary analyses submitted by stakeholders, a court would have a basis for finding unreasonableness. The same would be the case if the rulemaker were to assert that a chemical is not toxic (such that it need not be regulated) without engaging with submissions presenting evidence that the opposite is true.

Contrast these hypotheticals with a situation in which the stakeholder has merely set out a different view of the likely costs and benefits of a new rule, and now asserts that the rulemaker failed to show that its own views on this matter were more reliable. Or one in which the stakeholder merely asserts that the evidence the rulemaker relied upon was uncertain or open to multiple interpretations. If the court accepts arguments like these as a basis for striking down new rules, it goes beyond merely insisting on reasoned decision-making. It instead insists on a particular model of reasoned decision-making, in which the rulemaker must resolve away all sources of uncertainty before it can act. As Hilary Allen observes, the “practice of waiting until a problem has occurred before taking regulatory action is not a neutral approach. Instead, it stacks the deck in favor of [industries] who get to profit by generating risks that, if they come to fruition, will be borne primarily by the rest of society.”[149] For reasonableness review to be value neutral, it must leave space for the more democratic branches of government to make value judgments about what to do in uncertain situations, both through legislative delegations of power to rulemakers and their ongoing oversight of these rulemakers.

The ability to make value judgments would seem, however, to leave considerable room for arbitrariness. What should courts do if rulemakers err on the side of precaution in some situations, while waiting for more concrete evidence before intervening in others? Or if they depart from past practices in this vein as a result of a change of government? Or if they make value judgments that seem not to have a direct tie to their enabling legislation? The next subpart addresses these issues.

B. The Legitimacy of Value Judgments

Vavilov points to demonstrated expertise as the core justification for judicial deference to administrative action.[150] But as the preceding discussion illustrated, rulemaking is not just about the application of expertise. It inevitably involves making value judgments. To gain a clearer idea of why (and to what extent) courts should defer to these choices, we need to step away from the largely expertise-driven idea of deference in Vavilov and take a second look at the justifications for deference offered up in Katz Group.

Katz Group recognized that the legitimacy of value judgments reflected in rules flows not from rulemakers’ expertise, but from their congruence to value judgments reflected in enabling legislation.[151] In setting objectives to be pursued, enabling legislation reflects judgments about which stakeholders ought to benefit from new rules and at whose cost. Securities regulation, for example, is intended in large part to protect retail investors by imposing costs on the securities industry.[152] This means rulemakers are not free to exercise whatever value judgments they want—courts can and should assess whether these judgments are rooted in rulemakers’ enabling legislation. For example, if national security legislation demands substantial evidence that the benefits of new surveillance measures targeted at citizens will outweigh their costs, it would be unreasonable for the government to adopt measures rooted in precautionary analysis.

Of course, enabling legislation is prone to permit a broad range of different value judgments to be made.[153] For example, it might set out multiple, conflicting objectives that need to be balanced by the rulemaker.[154] Or it may not offer clarity as to how much evidence a rulemaker ought to collect before adopting new rules. But even this does not leave rulemakers with free rein to adopt whatever rules they want, on whatever basis they want. Whether rulemaking power is delegated to cabinet, a minister, or a regulatory agency, the rulemaker is connected to democratic components of government. Ministers are hired and fired from cabinet on the advice of the first minister, the heads of regulatory agencies are hired and fired by a minister or cabinet, and the cabinet as a whole depends on the confidence of the elected house of the legislature to remain in office.[155]

While it would be simplistic to view rulemakers as merely carrying out orders from their political principals in the executive or the legislature, work in the United States on similar relationships between rulemakers and the executive highlights how this accountability relationship constrains the kinds of value judgments reflected in new rules.[156] Rulemakers work to anticipate the broad policy priorities of the executive, often in reliance on political appointees, and map these priorities onto the more technical work they carry out.[157] Thus, rulemakers’ value judgments gain legitimacy from a combination of their congruence with those expressly laid out in the statute (which courts may enforce through judicial review), and the statute’s creation of accountability relationships between rulemakers and democratic branches of government (which the executive enforces through its hiring, firing, and oversight powers). City councils’ relationship with the executive is more complicated, of course, but these bodies have an even closer tether to democratic legitimacy: the relationship their enabling legislation creates between themselves and their electorate.

Recognizing the scope for value judgments in rulemaking would not render rules immune from reasonableness review. Let us return to the scenario introduced in the last subpart, in which a rulemaker makes different value judgments in adopting similar rules—taking a precautionary approach to regulating some financial products, and a wait-and-see approach to others, without offering some reasoned explanation for treating the products differently. Here, Vavilov’s emphasis on consistency in administrative decisions as a marker of reasonableness would seem to come into play.[158] A reviewing court could recognize, per Katz Group, that rulemaking involves value judgments, but still inquire, per Vavilov, into whether these value judgments are being made in a consistent, principled way. Otherwise, rulemakers might be able to cherry-pick value judgments issue-by-issue, a behaviour that would seem to reflect favouritism rather than good faith.

Of course, a change in government is prone to mean different value judgments will come to be reflected in rules. Reasonableness review should not impede these kinds of changes—otherwise, it would ignore the role enabling statutes give the executive (and, in the case of city council, voters) to guide the exercise of rulemaking powers. But it should prevent rulemakers from relying on purely “political reasons” for judgment—i.e., a statement that a rule was adopted or repealed merely because there was a change in government.[159] Rules still would need to “be connected to the relevant statutory criteria for judgment” and “justified by reference to the factual record.”[160]

C. The Need for Compromise

Because rulemaking involves value judgments on which reasonable people can disagree, rulemakers may need to reach compromises in the face of such disagreement. Members of municipal councils, for example, may need to bridge differing visions of the public interest in crafting by-laws. The provincial and territorial securities regulators must also reach agreement among one another before promulgating rules that are national in scope.[161] Further, rulemakers with oversight of different but related industries may find it necessary to coordinate among one another to tackle common problems.[162]

What is more, rulemakers may find it necessary to consider how the public might react to the rules they promulgate—not just because they are accountable to the public, but because public reaction is relevant to the effectiveness of these rules.[163] Rules meant to be comprehensive in scope, but that dramatically upset the status quo for the public, might be ignored or evaded.[164] Rulemakers might opt to avoid such dramatic action entirely or delay it until more consultation can take place (perhaps in the hope that more elaborate consultation procedures might increase public acceptance of the measures).[165] This means that a court engaged in reasonableness review ought to be willing to accept that rules will not constitute comprehensive solutions to the social problems they target, nor always reflect a coherent vision of the public interest. Rules may need to incorporate multiple visions of the public interest to secure agreement among different rulemakers. They may need to be narrowed in scope to account for public reactions to them. As such, rules should not be regarded as unreasonable simply because evidence suggests the rulemaker could have been justified in going further than it did when designing new rules, or that there are vulnerabilities in these rules that may require that they be revisited later on. Rather, as Katz Group recognized, courts should give rulemakers space to act incrementally.[166]

However, it is easy to see how action framed as incremental or as reflecting compromise might in fact reflect catering to interest groups. Rulemakers could discriminate against different segments of an industry, imposing burdens on one but not the other, and cite incrementalism as their basis for acting. Or, in the name of political compromise, they might rely on value judgments that do not correspond to those reflected in their enabling legislation. In conducting reasonableness review, courts can and should police against this conduct by reviewing rules in light of the legal and factual constraints on the rulemaker. Compromises in the face of uncertainty about how best to pursue the objectives of a statute, or how best to balance these objectives, are permissible. But rules that disregard the statute or the factual record are not.[167]

D. Encouraging Regulatory Lookback

Pervading uncertainty about the consequences of rules, and the resulting space for value judgments and compromise in rulemaking, means rules are unlikely to provide comprehensive and enduring solutions to social problems. Rather, as Katz Group explains, “an evolving awareness” of the drivers of social problems and regulated actors’ responses to new rules necessitate that rules evolve over time.[168] As such, it would seem unproductive to expect rulemakers to cram “all that could possibly be thought or dreamed about” the consequences of new rules into “single-shot, all-encompassing decision extravaganzas.”[169] But in exchange for being given scope to act in the face of uncertainty, rulemakers could be expected to do more to revisit and revise their decisions as they learn more about the consequences of these decisions.[170]

Reasonableness review can and should encourage this behaviour. This would give courts a pathway toward resolving another problem with giving rulemakers space to make value judgments in uncertain situations: what happens when rulemakers make judgments that reflect their enabling legislation and the facts as known at the time of adoption, but that turn out to be wrong? If legislation requires the periodic review of rules, this presumably would give stakeholders a direct mechanism for seeking reasonableness review of rules that have proven ineffective or harmful.[171] But even in the absence of these legislative measures, courts have options available other than flat-out “abnegation” to the executive’s value judgments and “tolerance for [the] arbitrariness” inherent in these judgments—the pathway Adrian Vermeule suggests courts learn to take when faced with uncertainty about the effects of new rules.[172]

Rather, when courts are presented with conflicting analyses of a rule’s likely costs and benefits at its adoption, they could give weight to a commitment by the rulemaker to monitor the effects of the rules it adopts and revisit them after a given period. If, by this point, these rules are failing to have their intended effects and there is no reason to expect this situation to change in the foreseeable future, the rules then could be subject to challenge if the rulemaker does not revise or repeal them.[173] Applying reasonableness review in this way would reflect the expectation that rulemakers will apply expertise in designing solutions to social problems, while also recognizing that this expertise will not allow for definitive responses to every concern or countervailing analysis provided by stakeholders before action is taken. These answers are only prone to become available after action is taken, at which point the rulemaker can apply its expertise to understand the effects its rules have had and whether these effects correspond with its predictions.

In the United States, these regulatory lookbacks are viewed as fostering sound policy, but implementing them has proved difficult—perhaps in part because of the US approach to reasonableness review.[174] By requiring rulemakers to give reasons that comprehensively respond to all vital comments on new rules, courts make these rules not only tougher to make but tougher to revise.[175] Rather than revisiting rules to see whether they had their intended effects, and thus opening themselves up to another round of intense stakeholder and judicial scrutiny, it may be rational for rulemakers to “steamroll their [initial] decisions through ... and then never look back.”[176]

On the approach to reasonableness review outlined above, by contrast, courts could work to encourage regulatory lookbacks. By recognizing that a rulemaker’s expertise likely will not allow it to resolve all sources of uncertainty before rules are made, and hence giving weight to commitments to review the effects of rules after implementation, reasonableness review can promote the kind of learning from experience that is part of a “more defensible ... conception of the administrative state” that reflects “aware[ness] of its own fallibility, that routinely invites challenges to its technical and political authority, and that continually responds to these challenges with reasons that are legible to the courts and to the public at large.”[177] Rather than inevitably leading to regulatory ossification, reasonableness review could serve to promote regulatory renewal.

The foregoing approach to reasonableness review is not a panacea. For one thing, periodic review of regulation may not always give rise to clear answers about the costs and benefits of new rules. Uncertainties present at the time of a rule’s adoption may persist in the years following its adoption. Technical and social developments may even introduce new sources of uncertainty.[178] And as the next Part highlights, substantive review of rulemaking is unlikely on its own to resolve the concerns raised at the start of this article about the quality of policymaking in Canada. It could, however, force courts to revisit other doctrinal issues relevant to this problem. And the analysis developed in this article may in turn have some bearing on these issues.

IV. frontiers of judicial review

If courts bring rulemaking within the scope of reasonableness review, there is reason to expect judicial attention to turn to at least two related issues. The first is whether interpretive guidance issued by administrators should also be subject to reasonableness review. If courts ignore this issue, policymakers looking to shield their decisions from judicial review would have considerable incentive to cloak new rules as guidance.[179] The second is whether rulemaking should be subject to review for procedural fairness as well as substantive reasonableness. If courts are to treat rulemaking and adjudication in broadly the same way for purposes of substantive review, their drawing a binary distinction between these decisions for purposes of procedural review seems difficult to justify from the standpoint of doctrinal coherence.

Both of these measures would bring opportunities and risks similar to those raised by reasonableness review of rulemaking—the potential for better-informed, higher-quality regulation, but also the risk that vested interests will be able to leverage procedural and other hurdles to frustrate regulation. Accordingly, this Part draws on the preceding discussion to sketch out some questions relevant to our evaluating these measures and possible responses. This is done not in a bid to provide complete answers to the questions of whether or how each of these measures should be adopted, but merely to try and move the conversation on these measures forward.

A. Substantive Review of Guidance

Guidance is ubiquitous. Any regulatory regime is bound to come with “guidelines, codes, manuals, circulars, directives, bulletins” and other materials produced by a range of different administrative authorities, with varying levels of importance.[180] While guidance is not technically binding on administrative decision-makers, unjustified departures from this guidance may be treated as unreasonable on judicial review.[181] And while guidance similarly is not binding on regulated actors, these actors may view compliance as preferable to a potentially costly fight with government over the proper interpretation of rules and legislation.[182] As such, for practical purposes, guidance “shapes a wide range of public decision-making.”[183] Nonetheless, while guidance can be challenged as unauthorized rulemaking to the extent its terms take on the trappings of a rule (e.g., by purporting to impose new obligations on regulated actors, or by conflicting with or purporting to modify another rule), true guidance falls outside the scope of judicial review.[184]

If rules and guidance both have practical impacts on stakeholder interests, should they not be subject to judicial review on the same standard? Should any reasonableness standard applied to rulemaking not also apply to guidance? As noted above, one advantage of such an approach would be to reduce administrators’ incentives to engage in avoidance behaviour, attempting to characterize rules as guidance to try and get around reasonableness review. This in turn should reduce the need for litigation over whether a decision is best characterized as rules or guidance.[185] But this advantage would come at the expense of restricting some of the flexibility said to be the key benefit of guidance. The lack of formality around the creation of guidance means it is also easy to update—administrators can work iteratively, creating and revising guidance in response to changing circumstances and new information about the effectiveness of various policy options.[186] If parties can immediately make submissions requesting the reconsideration of new guidance, and then seek judicial review of perceived failures to adequately act on these submissions, administrators may be reluctant to publish guidance without first gauging stakeholder reaction (e.g., through a notice-and-comment process). This may leave them without any practical tools for quickly responding to new problems or risks.

Rather than treat all guidance the same, it may be worthwhile to treat guidance documents differently depending on the level of formality around their adoption. Guidance adopted pursuant to more formal processes may be more amenable to reasonableness review, as this guidance seems unlikely to be highly iterative or fast-moving.[187] For example, securities commissions have express statutory powers to create policies that guide the application of rules, subject to completing a notice-and-comment process.[188] As such, these policies inevitably will take time to finalize even without the possibility of judicial review. However, guidance without this pedigree, such as notices or bulletins issued by civil servants without reference to any statutory authority or prescribed adoption process, could be treated as outside the scope of reasonableness review. Documents falling into this category would include the interpretation bulletins issued by tax authorities, which are neither subject to public comment nor contemplated by tax legislation.[189]

This is only one way of drawing such a dividing line, however. It may not even be the best one: commentary suggests the weight administrators give guidance bears little connection to the level of formality around its adoption.[190] Informal practices or staff guidance may have more influence on stakeholders’ practical interests than more formal, seemingly authoritative guidance. In that light, perhaps courts should subject guidance to reasonableness review depending on the level of weight it seems to have over administrative decision-making. Alternatively, it might be much simpler for courts to instruct administrators not to treat informal guidance as presumptively authoritative. Instead, perhaps administrators should be told to rely on this guidance only to the extent it is accompanied by reasons justifying the views it takes.[191]

Like the question of where to draw the line between reviewable and unreviewable guidance, the question of how to treat unreviewable guidance cannot be resolved here. But the prospect of subjecting rules to substantive review for reasonableness brings currency to the problem of whether and to what extent guidance should be subject to similar review. As such, it would seem prudent to at least start to frame a dialogue on this topic.

B. Rulemaking and Procedural Fairness

Adjudicators whose decisions affect “the rights, privileges or interests” of others owe a duty of procedural fairness to these affected parties.[192] Often, this duty entails giving prior notice and an opportunity to be heard before a decision is rendered, and giving reasons for the decision ultimately reached. But this need not always be the case: the duty is meant to be “flexible and variable,”[193] such that its implications for a particular administrative decision will be “appropriate to the statutory, institutional, and social context of the decision.”[194] Notwithstanding this flexibility, courts have refused to extend a duty of procedural fairness to rulemakers, citing concerns that such a duty would impede the rulemaking process.[195] This ongoing refusal has been criticized on at least two grounds. First, it leaves considerable room for rulemakers to make decisions without transparency and without justification (giving rise to risks of shirking, pandering, and catering to vested interests).[196] Second, it fails to recognize that the flexibility inherent in the duty gives courts scope to promote more accountable rulemaking without frustrating the rulemaking process.[197] In some instances, the duty could require that rules be subject to notice-and-comment, while in others it may simply require “transparency in the decision-making process” (e.g., by requiring the rulemaker to publish reasons or the record used in the rulemaking process).[198]

As suggested in Part II, subjecting rulemaking to reasonableness review could have procedural implications on its own. Even if rulemakers act without undertaking public consultation or reason-giving, they may be obligated to consider any submissions from affected stakeholders requesting that they reconsider their action, and a failure to act in accordance with such an obligation could be attacked through judicial review.[199] Accordingly, rulemakers may feel the need to develop procedures for responding to these requests, and to provide assurance that their decisions are rooted in the record and hence likely to survive judicial review.

Even if reasonableness review gives rise to these kinds of changes, however, this would not be enough to address concerns about the lack of procedural guardrails around rulemaking. Reasonableness review only elevates the quality of administrative decisions to the extent stakeholders can credibly threaten to seek review of decisions that fall short of this standard. And government has a number of levers available for reducing the credibility of this threat, especially from relatively vulnerable stakeholders. It can choose to consult some affected parties and not others,[200] perhaps in the hope that those left out of the consultative process will not organize to make submissions or force judicial review. It can neglect to publish submissions received from stakeholders, making it more difficult for affected parties to evaluate the reasonableness of the decisions taken without first launching litigation.[201] Regulated industries and other vested interests’ capacity to seek judicial review of decisions that impose costs on them gives rulemakers reason to take these stakeholders seriously. This is not necessarily the case for consumers, workers, and others—not to mention cost-bearers in the immigration, national security, and prison contexts—who are all far less well-placed to monitor or challenge rules affecting their interests.

If rulemaking and adjudication are to be placed on a similar footing for purposes of substantive review, doctrinal coherence could be viewed as demanding that they be subject to similar standards of procedural review, as well. What is more, given the concerns cited above about cloaking rules as guidance to avoid judicial review, similar guardrails should perhaps be considered with at least some forms of guidance. Of course, these guardrails would need to reflect the considerable uncertainty that attends the creation of rules and guidance, as well as the need for administrators to have scope to quickly respond to new problems. For example, notice-and-comment procedures may not be appropriate prerequisites for the adoption of public health measures in response to a pandemic.[202] Transparency about the evidence or value judgments relied upon in adopting these measures, as well as an opportunity to seek reconsideration of these measures after some time interval, may be more workable.[203] While the possibility of courts setting different procedures for rulemaking in different contexts creates uncertainties for rulemakers, they (and, for that matter, legislatures) can easily mitigate this uncertainty by imposing rulemaking procedures on their own. Though these procedures may not represent authoritative statements of the contents of a duty of procedural fairness in rulemaking, the case law makes clear that courts would accord significant weight to them.[204]

CONCLUSION

Vavilov set out to establish a “culture of justification,” in which administrative actors “demonstrate that their exercise of delegated public power can be ‘justified to citizens in terms of rationality and fairness.’”[205] Though resistance to the inclusion of rulemaking within this culture persists, the doctrinal reasons offered for this resistance seem highly vulnerable to attack. A more compelling basis for this resistance can be found in the potential consequences of extending Vavilov-style reasonableness review to rulemaking. This is especially the case where regulatory cost-bearers have the incentive and the resources to weaponize reasonableness review to frustrate regulation meant to serve the interests of consumers, workers, and other proxies for the public at large. Administrative law scholarship in the United States, where rulemaking has long been reviewable for reasonableness, suggests this concern ought to be taken seriously.

Rather than militate against the adoption of reasonableness review, however, this concern points instead toward a need to define reasonableness review in a way that more realistically reflects what rulemaking involves. Rulemakers, no matter how expert, can never be certain about the effects of their actions. It is therefore unrealistic to expect them to respond to every alternative scenario put forward by affected stakeholders, or to resolve away all sources of uncertainty about their stated basis for action. Rulemaking entails not only the application of expertise to problems, but the making of value judgments in response to uncertainty. Rulemakers should be entitled to make these judgments, provided they are congruent with those reflected in their enabling legislation. These judgments might be expressly stated in the legislation or implied through the creation of accountability relationships between rulemakers and the executive, the legislature, or voters.

But courts need not give rulemakers a blank cheque, under which the latter can act on uncertain evidence and never look back at the consequences of their actions. When faced with conflicting evidence of the possible consequences of a new rule, courts could give weight to a commitment by the rulemaker to review that rule after a given period to see whether it is achieving its intended effects. A failure to account for evidence of these effects could at that later point provide a basis for judicial review. Under this reframed approach to reasonableness review, courts would not only allow rulemakers to revisit and learn from their decisions but encourage it. This approach to reasonableness review will not resolve all concerns about the quality of policymaking, but it should at least help move the conversation forward.