Select Issues with Curial Deference on the Basis of Administrative Expertise A Behavioural Approach to Canadian Administrative Law’s “Prodigal Child” [1]

  • July 15, 2019

Giancarlo Mignardi, Queen's University

*Nota Bene: For the sake of brevity and simplicity, I have used the terms “administrator” or “administrative decision-maker” throughout the essay to describe those employed by the executive branch to support the functioning of the Canadian administrative state. I recognize that given the great variety of administrative bodies in Canada, these professionals may, in reality, go by very different titles.*


When should administrative decisions be afforded deference by a reviewing court? Why should deference be afforded to a given administrative decision? The apparently never-ending search for reliable answers to these questions has plagued Canadian lawyers, judges, academics, and students for decades. Currently, a fragile majority of the Supreme Court of Canada (SCC) has seemingly endorsed the notion that deference is typically warranted where a decision-maker possesses special expertise with respect to the subject matter of a decision. The possession of such special expertise is, moreover, typically presumed. Is this sound? Does it promote administrative decision-making that is effective, efficient, and gives due regard to the many rights and interests that often come to bear in a given decision?

The application of concepts and literature from behavioural economics to the law of judicial review for administrative action––in particular, to the more deferential reasonableness standard of review––can assist in answering some of the questions posed above. It is thus my intention in this paper to employ what behavioural economics has revealed about the fallibility of expert decision-making to the law surrounding the reasonableness standard of review in Canada and the assumptions underlying it. Such an inquiry will demonstrate that administrative expertise is not a wholly reliant basis upon which to defer to an expert administrative body, as these experts are prone to cognitive errors while making decisions. Given this, reasonableness review’s current functioning may require certain alterations in order to alleviate or minimize both the sources and the effects of expert decision-making fallibility.

To this end, in Part II, I delve further into discussing the current need for such an inquiry. After providing a summary of the current status of reasonableness review in administrative law in Part III, as well as a summary of what generally comprises the behavioural economic approach to analyzing human decision-making in Part IV, I move on to an assessment of the various cognitive errors to which experts are prone in Part V. In Part VI, I engage in a more specific discussion about how these cognitive errors and accompanying confusions regarding the scope of an expert’s expertise may lead to constitutional rights not being given their full expression when they are in play in a given administrative decision. Following this, in Part VII, I discuss why and how the provision of comprehensive reasons by the administrative decision-maker assists in minimizing the influence of some cognitive errors, and how the SCC’s current stance with respect to reason-giving is problematic. Finally, in Part VIII, I suggest some prescriptive measures that may assist in bettering the efficacy of expert decision-making, the most important among these suggested measures being a standard of judicial review that is, overall, more rigorous.


According to some, there is a relative dearth of behavioural analysis of administrative law.[2] While behavioral economics has revolutionized legal scholarship in many areas of law, this is not so in administrative law.[3] Indeed, there is, to date, no self-identifying field of “behavioural administrative law and economics”.[4] This is especially true for Canadian administrative law. As will be demonstrated, this is unfortunate, particularly given the extent to which foundational administrative law doctrine in Canada has centrally relied upon views about the administrate state and administrative expertise that directly implicate––or at least should be directly implicating––behavioural issues.

Canadian administrative law’s standard of review analysis––particularly, the reasonableness standard of review––has been the subject of increasing negative commentary with respect to its efficacy in balancing legislative supremacy and the rule of law. More importantly for the purposes of this paper, the high level of deference reasonableness review affords to the decisions of administrative decisionmakers in the overwhelming majority of cases coming before the SCC has been grounded upon the notion of tribunal/agency expertise. This connection between deference and expertise––the latter being the putative justification for the former––suggests that a behavioural analysis of tribunal/agency decision-making is well warranted. If it is the case that the quality of administrative expertise is lacking, then the abovementioned connection seems somewhat spurious, and the currently endorsed notion, in Canadian administrative law, that expertise inheres in a tribunal/agency itself[5] seems somewhat problematic. A behavioural inquiry, in short, may allow for a deeper dive into this idea of expertise and what it can entail.

It could even be said that such an inquiry could not conceivably be timelier. The end of 2018 is a significant time for Canadian administrative law, as, just a decade after Dunsmuir[6] was decided, the SCC has granted leave to appeal to a trilogy[7] of administrative law cases and, in so doing, has specifically indicated, in granting leave to two of those cases,[8] that it is:

… of the view that these appeals provide an opportunity to consider the nature and scope of judicial review of administrative action, as addressed in Dunsmuir … and subsequent cases. To that end, the appellant and respondent are invited to devote a substantial part of their written and oral submissions on the appeal to the question of standard of review, and shall be allowed to file and serve a factum on appeal of at most 45 pages.[9]

To reinforce the notion that this could provide an opportunity for the SCC to consider implementing fundamental changes to the common law of judicial review, 27 interveners were involved in the consolidated SCC appeals, and the Court has appointed numerous amici curiae to assist it by means of written and oral submissions.[10] In other words, facilitating discussions that force stakeholders to probe and potentially rethink or add nuance to, inter alia, the theoretical, empirical, and policy justifications undergirding the current practice of curial deference is warranted right now because substantial changes and improvements to the standard of review may indeed be realizable very soon (i.e. at some point in 2019 when the decisions are released for this abovementioned trilogy). Granted, many commentators are less than enthusiastic. Daly, for one, seems to be in a fairly overt state of pessimism:

What can we expect in these appeals? As a general rule, there is an inverse correlation between one’s level of anticipation of an administrative law decision and the importance of the decision. Put another way: the more excited administrative law aficionados are beforehand, the more disappointed they will be afterwards.[11]

Daly does have a point here––one that will be explored more in Part VIII, below––in that an invitation to have Dunsmuir reconsidered will not ultimately mean that all those commentators who have been frustrated with the jurisprudence of the past decade (and beyond!) will be able to see brighter days ahead of them as a result of this trilogy, as this will almost certainly not be the case. This should not, however, detract from the fact that Canadian administrative law “aficionados” nevertheless find themselves at a crucial jurisprudential juncture––one that should call them to try assisting in the facilitation and dissemination of ideas about the way forward that current and future generations of stakeholders may consider.

Therefore, while a comprehensive discussion regarding what issues are before the SCC in this trilogy, as well as how they could or should be resolved, is largely beyond the scope of this paper, it is nevertheless important to acknowledge that the critique of select foundational administrative law principles (and cases) found below comes at a unique time of potentially great flux in the field. As a result of this, one objective within these circumstances is for this paper to form part of the contemporary writings––both before the SCC in the form of submissions and outside the court in the form of academic and professional commentary (and consternation)––seeking to communicate that any major change to the standard of review, ideally, should not only consider where Dunsmuir could be “tweaked”, but should also acknowledge that achieving a better standard of review necessitates a comprehensive and (hopefully) empirically-informed reappraisal of foundational principles––curial deference on the basis of expertise being one of them. Put simply, it is hoped that the SCC does more than institute “tweaks”, and I briefly address this further below in Part VIII.


While it is certainly not my objective to discuss every controversial issue with the current state of Canadian administrative law and, in particular, the reasonableness standard of review found therein, I wish to introduce some significant components of and assumptions underlying reasonableness review that are important for the purposes of this inquiry. To begin, I wish to clarify that the plainest reason for the particular focus on reasonableness here is that, as Macklin has noted with respect to the standard of judicial review employed in Canadian administrative law, “[s]uperfically, it seems simple: it’s (almost) all reasonableness, (almost) all the time”.[12]

As Dunsmuir contemplated and as the post-Dunsmuir jurisprudence has confirmed, reasonableness is a deferential standard:

Deference in the context of the reasonableness standard therefore implies that courts will give due consideration to the determinations of decision makers. As Mullan explains, a policy of deference “recognizes the reality that, in many instances, those working day to day in the implementation of frequently complex administrative schemes have or will develop a considerable degree of expertise … to the imperatives and nuances of the legislative regime … ”. In short, deference requires respect for the legislative choices to leave some matters in the hands of administrative decision makers, for the processes and determinations that draw on particular expertise and experiences … .[13]

As the above demonstrates, not only is reasonableness a deferential standard; it is a deferential standard rooted, inter alia, on recognition of and respect for administrative expertise.[14]

Importantly, for several reasons, some of which are beyond the scope of this paper––issues of legislative intent and statutory interpretation––reviewing courts in Canada generally do not engage in a factual inquiry to assess the level of administrative expertise, nor have they permitted reasonableness’ levels of deference to vary depending on the assessment of the quantity or quality of expertise. These are questions that more or less go unasked. As McLachlin CJC has stated:

Relative expertise can arise from a number of sources … . The composition of an administrative body might endow it with knowledge uniquely suited to the question put before it and deference might, therefore, be called for … . Similarly, an administrative body might be so habitually called upon to make findings of fact in a distinctive legislative context that it can be said to have gained a measure of relative institutional expertise … [emphasis added].[15]

Indeed, as Van Harten et al. note, while the factor of statutory purpose has been integral to the SCC’s approach since CUPE[16], since the mid-1990s, expertise has come to the fore as a primary consideration in calibrating deference.[17] According to some, cases such as Pushpanathan[18] and Southam[19] demonstrated that expertise has become the most important of the factors to be considered in establishing the standard of review.[20] Yet, as Cameron observes, the legal criteria that courts elaborated for determining quality of expertise were largely formal.[21]

How formal? For one, Dunsmuir had initially said that deference will usually result “where a tribunal is interpreting its home statute …”[22]. Yet, within a few years, this statement had been elevated to a presumption.[23] Why? The SCC consistently offers legislative delegation and expertise as the significant component parts of the overall rationale for curial deference. [24] Regarding deference, consider, below, how far certain SCC justices have gone to “respect” it:

Expertise arises from the specialization and functions of administrative tribunals … . However, as with judges, expertise is not a matter of the qualifications or experience of any particular tribunal member. Rather, expertise is something that inheres in a tribunal itself as an institution [emphasis added].[25]

For the purposes of this discussion, the importance of the SCC’s abovementioned stance cannot, in my respectful opinion, be overstated. It means that, under reasonableness review, administrative decision-makers need not––conceivably ever––demonstrate or evince a requisite quality of expertise.[26] Putting aside the logical fallaciousness of functionally proclaiming that what is true of the whole (the tribunal) is true of its parts (individual decisionmakers), the notion, confidently adopted by the SCC, that administrative bodies necessarily possess the requisite quality of expertise seems particularly bold and suspect. If the veracity of such a notion were revealed to be at least partially lacking, it would seem as if a vital underlying justification for a deferential standard of review receives a major blow. As suggested above, a behavioural analysis of tribunal expertise allows for this notion to be put to the test. Before this is done, though, I first turn to a brief discussion of fascinating field of behavioural economics and how its tools may provide some needed assistance in this matter.


As is well known, the economic analysis of law has traditionally assumed that people are rational and self-interested maximizers of their own utility.[27] In the 1980s, though, certain studies explored the ramifications of cognitive and behavioural studies for the economic analysis of law and legal policymaking.[28] These studies produced such influential effects and spurred so much commentary that, by 1999, Cass Sunstein was calling the outpouring of growing work in behavioural law and economics “a flood”.[29] Sunstein, an influential voice in this field himself, would go on to advise the Obama administration on the formulation of more behaviourally informed regulations as Administrator of the White House Office of Information and Regulatory Affairs.[30] Eric Ip notes how some commentators have gone so far as to laud behavioural economics of law as the economic analysis of law.[31] But, before attempting to apply it to administrative law, let us explore behavioural economics itself.

Behavioural economics is, before anything else, concerned with undertaking descriptive analyses of economic decision-making on the ground. This is why it employs cognitive and psychological concepts to economic decision-making––rather than explaining how humans should think, judge and decide, behavioural economics explores how humans actually think, judge, and decide. Indeed, it has been observed that there is now a “distinction between normative theory, which describes the judgments that should be made, and what people actually do. People are systematically biased in these cases”.[32] From this descriptive inquiry has come the crucial observations that humans are not rational maximizers of their own utility. Instead, because of limited levels of, inter alia, time, attention, experience, expertise, or interest, humans sometimes make decisions using mental “shortcuts”––heuristics and biases––that, while quick and efficient, expose us to greater levels of irrational, fallible, incomplete or biased decision-making.

In Thinking Fast and Slow[33], Daniel Kahneman offers a succinct synopsis regarding how both quick and more contemplative decision-making occur by employing the language of “System 1” and “System 2”. He notes that, as has become well known, the traditional System 1 “operates automatically and quickly, with little or no effort and no sense of voluntary control”, while System 2, if awoken, “allocates attention to the effortful mental activities that demand it, including complex computations.”[34] Kahneman posits throughout his book that most of our daily decisions are made via System 1, which seems to constantly be inputting past experiences into the framework of certain biases and heuristics in order to assist humans in making snap, intuitive judgments about the world. Most of the time, these judgments are both helpful and correct. Yet, he warns that, while the operations of System 1 can generate complex patterns of ideas, “only the slower System 2 can construct thoughts in an orderly series of steps.”[35] In other words, biases and heuristics, while widely (and even involuntarily) employed, can and do lead us astray. This becomes all the more true if the impressions, intuitions, intentions, and feelings of System 1 are consciously endorsed by System 2, whereby “impressions and intuitions turn into beliefs, and impulses turn into voluntary actions”.[36]

The pervasiveness of efficient, but sometimes off-the-mark, System 1 decision-making does not, however, mean that human behaviour is rendered unpredictable or arbitrary. Along these lines, Sunstein reminds readers that bias- or heuristic-related departures from standard economic assumptions are often both systematic and predictable, thereby making them a legitimate basis for tracing and predicting human behaviour.[37] Indeed, much of behavioural economics is grounded upon this very basis, and the results of countless empirical studies in the field confirm the predictable nature of the above (in laboratory settings, at the very least).

These mental shortcuts operate almost systematically, such that human beings are often described as being boundedly rational. Moreover, the shortcuts come in many forms. As Sunstein observes:

Biases fall in various categories; they include hindsight bias, optimistic bias, and extremeness aversion. Efforts to economize on decision costs are responsible for rules of thumb, or heuristics. … [Heuristics] reduce the costs of making decisions, but they may not be fully rational if people could make far fewer errors by acting as good Bayesians.[38]

Why the concern, here, for biases, heuristics, and bounded rationality? To put it simply, they have been found to affect both lay and expert decision-makers alike,[39] such that those in government may be engaging in faulty decision-making.[40] For instance, in the context of the application of prospect theory, most studies have found that professionals are just as susceptible as laypersons to framing effects, displaying risk aversion when outcomes were framed as gains, and risk-seeking when the same outcomes were framed as losses”.[41] Sunstein has noted that because of the fact that cognitive errors do tend to be made at the administrative level, it seems that “the cognitively biases lead the cognitively biased.”[42] Insofar as this is true, it implicates issues surrounding the reasonableness standard of review. Should deference be based upon administrative expertise? Are experts reliable decision-makers? Do they succumb to biases and employ heuristics in problematic ways? If so, what does this say about reasonableness? Are there solutions? I now turn to some of these questions.


I begin here by noting that the following will discuss only some of the most salient “mental shortcuts” that could potentially be affecting expert administrative decision-making. Such an analysis, especially without the assistance of an accompanying empirical inquiry, is necessarily ad hoc. As such, there may very well be other biases and heuristics affecting expert decision-making and, indeed, inquiries into this should ideally be undertaken in a Canadian administrative setting in the near future.

As described above, administrative expertise seems to now be presumed within the context of judicial review,[43] even where explicit markers of expertise may not be present.[44] As a starting point, it is interesting that, even from outside the realm of behavioural considerations, this state of affairs has been attacked as problematic. For instance, Jordan highlights how other commentators have taken issue with a “broad application of the reasonableness standard based on the blind faith that all administrative decision makers are experts”.[45] A behavioural analysis will shed light on the extent to which this is truly “blind faith”. As will become clear, the issue is not so much that administrators are not experts, but rather, that experts seem to be much more fallible than the SCC has been willing to recognize.

a. Administrative Decisionmakers are Experts: A Concession

Most administrators are often undeniably experts. As Rachlinski and Farina make clear, “[t]he only real experts in the government are found in administrative agencies”.[46] Indeed, for many decades, one central motivation and justification for the very proliferation of the administrative state in Canada has been the desire to have field experts make policy and adjudicate in increasingly specialized areas of government. It may even be said quite confidently that, not only are administrators experts in a given policy area writ large, but they are typically also experts in the “ins and outs” of the types of adjudicative decisions made within that policy area. Somewhat like a trial judge who partakes in “ ‘hands-on’ engagement with the parties, the witnesses, and the facts”,[47] an administrator is called upon––often daily, depending on the institution––to engage with, inter alia, the parties, records, professionals, stakeholders, technology, and literature directly related to their board/agency/tribunal. Insofar as this is true, then administrators should not simply be shunned as non-experts. The issue, again, is not whether they are experts but, rather, whether the quality of their expertise should be presumed to be reliable.

Not only are administrators experts, but they are experts that society cannot do without. The legislature is not equipped with the time, resources, personnel, or even constitutional authority to apply and execute policy. As for the judiciary, most superior court judges are expected to be generalists who are “ready for anything”,[48] and are already operating under immense time and resource constraints doing just that. Administrators and the knowledge they possess, then, are vital.

b. An Expert’s Expertise is not Infallible: A (Behavioural) Qualification

Specialized knowledge, however, is not enough.[49] Administrators, as it were, possess a System 1 and a System 2 like everyone else, potentially rendering them susceptible to fallibility in making decisions and applying their expertise. Insofar as regulators suffer from the biases that plague laypeople, they are likely to use heuristics to estimate optimal policy choices or adjudicative decisions, and to err while doing so.[50] Moreover, Rachlinski and Farina postulate that the extra or special information that experts possess––the sine qua non of very notion of expertise––may even feed certain prevalent cognitive illusions[51] due to the fact that, “[i]f experts rely on heuristics that lead them astray, the extra knowledge they bring to a problem may become useless [or] even counterproductive”.[52]

Cooper and Kovacic also provide some insight here. They ask: “Is it reasonable to assume that because regulators are ‘experts’ and face similar problems repeatedly they will be able to make unbiased policy decisions?”.[53] While they believe this to be unlikely, they do suggest that expert decision-making can improve if those repeated decisions are followed by constant feedback, but that “as the link between a decision and feedback weakens, and ability to learn diminishes”.[54] Kahneman has similarly remarked that experience (repetition) and feedback constitute the two basic conditions for acquiring expertise,[55] and that “[w]hether professionals have a chance to develop intuitive expertise depends essentially on the quality and speed of feedback, as well as on sufficient opportunity to practice”.[56] This, then, not only means that environments providing infrequent feedback do not alleviate expert fallibility, but that problems also arise where the environment itself is one in which it may often take a long time to see the results of a decision.[57] Zamir and Teichman have suggested something similar.[58] Rachlinski and Farina, as well, note:

Experts have two opportunities for developing such decisionmaking competence that are typically unavailable to lay decisionmakers: experience and training. … Experience accompanied by feedback allows experts to identify situations in which they are using inappropriate heuristics or are trapped by misleading schema.[59]

The question then ostensibly becomes whether administrators at the countless agencies/tribunals/boards across Canada possess the requisite level of experience and appropriate levels of feedback. Assuming the experience is there, an issue arises with feedback. It is a nebulous term, and this paper does not include any empirical data with which feedback may be analyzed. With that said, one tentative observation that can be made is that, traditionally, judicial review has been a means through which the mechanics of the Canadian Constitution[60] have provided for the opportunity for the administrative state to receive feedback regarding whether its decision-making has reasonably (or correctly) conformed to statutory objectives, the rule of law, and the interests and rights implicated therein. Insofar as the judicial standard of review becomes too deferential, it therefore means that at least one strong source of feedback is not adequately available. In other words, if deference to administrators within the context of Canadian administrative law has truly become, as some have suggested, “deference as abdication”[61], then administrators may be lacking some important feedback. Without it, the above suggests that they will be more prone to err.  

c. Select Biases/Heuristics to which Experts are Prone[62]

1. Status Quo Bias

A likely bias toward the maintenance of the status quo is one issue in the administrative state.[63] Cooper and Kovacic describe it as a sort of  “inertia to maintain a current course of action rather than take new action that would increase expected utility”.[64] This bias has traditionally been explained via the combined presence of loss aversion and related endowment effects among administrators––the former creating an impression that losses from the status quo are felt more pronouncedly than any gains therefrom, and the latter further ossifying the desire to maintain the status quo and ascribe it greater value than it otherwise would (should) have. Indeed, as Eric Ip posits with respect to government regulators:

Loss aversion plus the endowment effect, both of which underpin status quo bias, explain how agencies’ initial set of implementing regulations for a regulatory statute commits regulatory policy to a particular path, dependency upon which, a by-product of status quo bias, discourages any tendency toward changing over to better but divergent policy paths.[65]

With respect to potential implications of the above, it is my view that a status quo bias, if extant, would cause the greatest problems in the policy sphere. An administrative agency that stubbornly clings onto the status quo in the face of institutional or statutory change would seem to frustrate the legislator’s intentions more than justiciable rights, especially in instances of, say, government or institutional change. That said, this bias could have an effect on the law when, for example, an administrator lags in implementing the feedback received from recent court decisions impacting its work. In any case, a failure to faithfully or adequately adopt and implement a given policy is still potentially unreasonable and, if identified, should not be afforded deference by a court. Why? In such cases, the administrator may be proceeding to make decisions that satisfy an arbitrary, pre-existing standard,[66] which becomes all the more arbitrary if it is outdated. This potential undermining of current policies, regulations, or objectives in favour of the status quo is concurrently undermining legislative supremacy, and should not be afforded deference.

2. Confirmation Bias

Administrators are not above succumbing to the confirmation bias.[67] This bias particularly reveals itself when administrators, attempting to dedicatedly and stalwartly further the objectives of their institution, make decisions wherein less attention is paid to information that contradicts their interests in a given matter, and more (or exclusive) attention is paid to information that protects or forwards these interests.[68] This is problematic, particularly in decision-making situations in which the rights or interests of others may impede administrative efficiency or objectives. Indeed, insofar as it may cause administrators to “filter new information so as to confirm the initial wisdom, without genuinely examining its merits”, the presence of a strong confirmation bias may provide fertile ground for the undue or unreasonable disregard for the substantive rights of others.[69] On the policy side of things, the bias “may predispose regulators to conserve existing regulations without critically evaluating their worth against relevant evidence, which may be complex and subject to multiple inconsistent interpretation”.[70] This may lead to inefficient decision-making that is out of step with the evidence as a whole, especially, for instance, when evidence contradicting agency/tribunal beliefs is construed as supporting it.[71]

There are two especially distinct problems with the confirmation bias with respect to both policy and legal issues. First, the potential pervasiveness of the confirmation bias is a large cause for concern. There is a risk of it being at play across all administrative decisions. This does not necessarily lead to unreasonable or inefficient decisions in all cases, but this could be an eventually at least some of the time. Second, the bias is difficult to eradicate. Simply feeding the administrator with more information or evidence is not enough, as these new materials may be subject to the bias themselves.[72] Overall, these problems may lend further support to the notion that a standard of review that is too deferential risks permitting this bias to go unchecked.

3. Propensity to be Overconfident

The presence of a heuristic of overconfidence while making decisions is a bias that is particularly pronounced for experts. Quoting Dale Griffin and Amos Tversky, Choi and Pritchard wittily remark at how experts are “often wrong but rarely in doubt”.[73] While expertise and specialization are usually cited as the key advantages of having professional or technocratic administrators, these groups tend to “overestimate their own abilities in fields where they think they have or have access to cutting-edge knowledge, blind to the advisability of consulting a fuller body of knowledge”.[74] Unfortunately, then,  experts may not be getting it right as often as they think.[75] As such, when this heuristic is at play alongside a sprightly confirmation bias, the expert runs the risk of totally disregarding more efficient or effective alternatives to the chosen course of action.[76]

In the context of more menial or workaday tasks, expert overconfidence may also promote too much cognitive ease. As Kahnemen describes, “[w]hen you are in a state of cognitive ease, you are probably in a good mood, like what you see, believe what you hear, trust your intuitions, and feel that the current situation is comfortably familiar. You are likely to be relatively casual and superficial in your thinking” [emphasis added].[77] The problem here is that the confidence or level of ease experts possess when they believe that their existing intuitions are applicable is not a reliable guide to their validity.[78]

The heuristic is a problem in the context of administrative law because it directly calls into question the reliability of expert decision-making. This is perhaps especially grave given that reviewing courts are composed of generalist judges who may not, themselves, be able to trace where the expert went wrong in complex scenarios. This latter eventuality should not, however, dissuade courts from probing the evidentiary record to identify objectionable instances of expert neglect for or indifference toward vital data. Indeed, as Ip has suggested, this should be done because too much deference to the experts “may be seen through the lens of [behavioural law and economics] to promote [greater] overconfidence and confirmation bias, among other cognitive problems”.[79] The judicial review of administrative action should not be promoting such things.

4. Availability Heuristic

The availability heuristic is one that, if present, would cause experts to ground their conclusions about a matter not on comprehensive evidence or representative data but, rather, on the existence of similar events that come to mind quickly.[80] These events may be anecdotal or unrepresentative outliers, but their recency or vividness or coverage in the media might lead them to weigh heavily on the administrator’s mind.

While the availability heuristic would presumably create more problems in the policy sphere––wherein administrators and regulators are constantly working under time and resource constrains to decide what the public interest is in a given case and how it may be realized––the heuristic may serve as a support for decisions that impact legal rights, as well. Consider the  Baker[81] case, where the evidentiary record before the SCC showed that, in determining whether Ms. Baker should be admitted into Canada permanently on humanitarian or compassionate (“H&C”) grounds, Officer Lorenz wrote, inter alia, the following:

[Ms. Baker] is a paranoid schizophrenic and on welfare. She has no qualifications other than as a domestic. She has FOUR CHILDREN IN JAMAICA AND ANOTHER FOUR BORN HERE. She will, of course, be a tremendous strain on our social welfare systems for (probably) the rest of her life. There are no H&C factors other than her FOUR CANADIAN-BORN CHILDREN. Do we let her stay because of that? I am of the opinion that Canada can no longer afford this type of generosity.[82] [Capitalization in original]

Did Officer Lorenz have sufficient and objective evidence or statistics to support the notions: that Ms. Baker would “of course” be a “tremendous” strain on Canada’s welfare system for the rest of her life?; or that Canada can no longer afford “this type” of generosity? Perhaps he did, but given the sheer boldness of the claims themselves, it is more likely that he did not. While this is ultimately mere speculation, what seems more likely is that Officer Lorenz had other vivid (and anecdotal) situations popping up in his mind surrounding newcomers to Canada being accused of causing strains to the Canadian economy. To the extent that this was the case, Officer Lorenz may have been plagued by the availability heuristic. A reviewing court, in similar instances, should be making clear to the administrator that conclusions arrived at via anecdotal or unrepresentative information may be unreasonable (or incorrect).

5. Groupthink

Finally, groupthink may also pervade administrative bodies. Groupthink emerges when individuals stalwartly identify themselves with the larger group of which they are a part and concurrently accept its objectives uncritically due to the individuals’ perceived membership in it.[83] Once this commitment is made, members will downplay views that are inconsistent with the group’s own views.[84] Government agencies can be home to a culture of groupthink, thereby undermining the critical thought, creativity, and open-mindedness of individual members who, instead, feel compelled self-censor and defer to the group.[85] In such settings, shared and common information “is more likely to be brought up, and repeatedly so, so it is likely to have a greater impact”.[86] The problem is that the shared and common information may itself be skewed or lacking.

On its face, groupthink does not seem to be a cause for concern provided that the organizational culture is open, dynamic, diverse, and strives for impartiality. But these may become lofty provisos, especially when a particular political stances, ideologies, or “ways of doing things” has become engrained in the institutional culture. In such scenarios, amended legislation, changing governments, new personnel, or impactful court decisions that challenge the institutional culture may be partially ignored or undermined. If this were to happen, both legislative intentions and relevant jurisprudence may be at risk of not being wholeheartedly considered by a given administrative body.


After analyzing some of the prominent biases and heuristics to which experts plying their trade within Canada’s administrative state may be susceptible and the general ways in which this undermines some of the assumptions underlying the deferential standard of reasonableness review, I now turn to a more specific discussion of the employment of reasonableness review­­––and the affording of deference––in the context of Charter[87] rights being implicated during the course of an administrative body’s decision-making. When this is the case, does reliance on and deference to administrative expertise––something that the above has demonstrated to be quite fallible––offer the right amount of protection to constitutional rights, the likes of which should be residing at the very apex of Canada’s legal order? An approach grounded in the considerations of behavioural economics suggests not. Given this, a reconceptualization of judicial review––or, oddly enough, a more stringent adherence to the text of Dunsmuir[88]–should be at least considered in cases where constitutional rights are implicated.

a. DorĂ©’s Directive: Reasonableness with a Pinch of Proportionality

In 2008, in Dunsmuir, the SCC affirmed that constitutional issues “are necessarily subject to correctness review because of the unique role of s. 96 courts as interpreters of the Constitution”.[89] Following this, though, there remained a lack of clarity with respect to whether a rigorous proportionality test would also apply once a probing correctness standard was triggered to review how the constitutional right had been balanced by the administrator. The SCC itself struggled to agree on what to do.[90] A full discussion about the Court’s journey in choosing the applicable analysis here is somewhat beyond the scope of this paper, but suffice it to say that it seems as if neither a correctness review nor a proportionality test currently apply to administrative decisions that were required to interact with and balance the constitutional rights of a party to a tribunal/agency/board. As Macklin succinctly notes, [c]urrently, a wobbly majority endorses deference to Charter determinations conducted in the course of discretionary decisions. This significantly retracts the scope of correctness review for constitutional questions.”[91]

The above reflects the SCC’s position unanimously communicated in the 2012 DorĂ©[92] decision. Here, the Court not only made clear that administrative decisions––state action that is purportedly (and perhaps oddly) distinct from laws and regulations in the books––were subject to a (deferential) reasonableness review in cases where the decisions engaged Charter rights or “values”. Justice Abella, writing for the Court in DorĂ©, was of the stance that this arrangement struck the right balance between granting administrators with the respect and autonomy they need to pursue statutory objectives on the one hand, and rigorously protecting rights on the other. For one, she noted that empowering administrators to interpret and apply constitutional values “allows the Charter to ‘nurture’ administrative law, by emphasizing that Charter values infuse the inquiry”.[93] Justice Abella was also of the belief that rights would receive due care under reasonableness because “the reasonableness analysis is [in such cases] one that centres on proportionality, that is, on ensuring that the decision interferes with the relevant Charter guarantee no more than is necessary given the statutory objectives”.[94] Finally, DorĂ© sought to keep the sacrosanct entitlement to deference alive on the basis of, of course, expertise:

When Charter values are applied to an individual administrative decision, they are being applied in relation to a particular set of facts. Dunsmuir tells us this should attract deference.[95]

Deference is still justified on the basis of the decision-maker’s expertise and its proximity to the facts of the case. Even where Charter values are involved, the administrative decision-maker will generally be in the best position to consider the impact of the relevant Charter values on the specific facts of the case [emphasis added].[96]

In light of the behavioural considerations hitherto discussed in this paper, one question that naturally arises is: do administrative experts truly possess a “distinct advantage”[97] when they attempt to apply the Charter to a specific set of facts? As will be seen, answering this does not involve calling into question the existence of administrative expertise but, rather, the quality and scope of it. If it is the case that the quality and scope of expertise here do not allow for Charter rights to be considered to their fullest extent due to behavioural biases and heuristics, is an overly deferential standard a problem? Can administrative decision-making be calibrated in a way that “centres” on proportionality? If not, could this all not lead to decisions that do not grant the rights of individuals their fullest expression, and a corresponding abdication the “unique role of s. 96 courts as interpreters of the Constitution”?

b. Problems with DorĂ©’s Directive: A Behavioural Analysis

As abovementioned, rather than adapt the Oakes proportionality test to the exercise of an administrative decision, the court proposed a reasonableness analysis that balances “the severity of the interference of the Charter  protection with the statutory objectives”.[98] I will posit, here, that behavioural economics considerations seem to suggest that the problems stemming from this analytical framework are twofold. Firstly, the biases and heuristics explored above, if applied to this context, would lead to decision-making results that do not, in fact, give due or reasonable regard for rights. This is more of an issue with the quality of expertise in this context. Secondly, because expertise only extends so far, it seems to be the case that administrative experts in a certain policy domain are not necessarily constitutional experts within that same domain. This latter problem is more of an issue with the scope of the expertise in this context.

1. Quality of Expertise: Biases, Heuristics, and how Statutory Objectives Trump Rights

To state the probable obvious, the notion that it may be awkward to an administrative body’s balancing of a constitutional right while it also considers a concoction of other interests and objectives is not new. Consider, for instance, the words of the British Columbia Court of Appeal (BCCA) in its 2016 Trinity Western University decision:

We would observe … that many tribunals have limited contact with the Charter and may have considerable difficulty interpreting it. There is also a real possibility that a tribunal’s preoccupation with its own statutory regime will lead it to value the statutory objectives of that regime too highly against Charter values. As well, it is important to recognize that administrative tribunals do not enjoy the same independence that judges do. … [It] may be vulnerable to public or governmental pressure, and may find it difficult to give the Charter rights of unpopular persons or groups sufficient weight when balancing them against statutory objectives.[99]

Wildeman echoes the concerns of the BCCA when she remarks that “government and state-appointed actors may be too embedded in sector-specific values or majoritarian interests to warrant deference on matters involving fundamental human rights”.[100] Ponomarenko does very well to add to these concerns, observing that, whereas reasonableness review essentially holds deference as being of presumptive importance,[101] a more Oakes-like proportionality review would tend to grant presumptive importance to the protection of rights, demanding limitations on rights are only acceptable insofar as they have been demonstrably justified.[102]

Viewing the above from a behavioural perspective seems to lend support to the concerns of these writers. Recall, for one, the presence of a confirmation bias during the course of administrative decision-making could mean that, insofar as a statutory objective is categorized as the primary consideration in the mind of an administrator, that information undermining the efficient pursuit of that objective may not be given the time and acceptance that it deserves. Indeed, as Ip has noted with respect to regulators, once they “have committed themselves to a particular policy position, they become prone to filter new information so as to confirm the initial wisdom, without genuinely examining its merits”.[103] Similarly, Cooper and Kovacic believe that “[f]lawed heuristics … are likely to lead regulators to adopt policies closer to the preferences of political overseers than they would otherwise.”[104] This does not seem to bode well for parties before a tribunal whose constitutional rights are in play, unless protecting a right does nothing to hinder the pursuit of the statutory objective. But this is certainly not always the case. Recall, furthermore, that the presence of an overconfidence heuristic may mean that administrative experts may truly believe that their experience and intuitions are serving them well even while balancing constitutional rights, and so they may not even intend to or be aware of other potential ways to have given rights their full expression in a given case. Finally, recall that a preference for the status quo, greatly undergirded by an aversion to loss, may lead administrators to believe that a method of protecting a right that seems too onerous on administrative time or resources may be labelled as “too much”. Again, this does not bode very well.

2. Scope of Expertise: Experts do not, in fact, Know Everything

Perhaps even more fundamentally, there seems to be a problem with the SCC’s assumption, on full display in DorĂ©, that an administrator’s expertise will serve it well when it is called upon to balance Charter rights. Here, too, concerns from legal academics abound. Macklin, for one, believes “[t]here is simply no basis for a presumption that a decision-maker’s ‘field expertise’ … equips the decision-maker to manage all aspects of a Charter analysis”.[105] This makes intuitive sense. For example, a telecommunications expert may have a wealth of knowledge with respect to telecom-related statutes, telecom infrastructure, and the many interests of telecom stakeholders. It does not follow that this same person will know whether or how to grant a consumer’s freedom of expression its full, well, expression across various telecom media.

Here, the behavioural lens provide insight. As Kahnman notes, “[e]xpertise is not a single skill; it is a collection of skills, and the same professional may be highly expert in some of the tasks in her domain while remaining novice in others”.[106] Unfortunately, though, experts are not necessarily aware of the limits of their expertise; that is, potentially due to overconfidence, they may not have learned to identify cases in which their expert intuitions may betray them.[107] If this is true, how can it be reconciled with the SCC-endorsed claim that “the administrative decision-maker will generally be in the best position to consider the impact of the relevant Charter values on the specific facts of the case”?[108] In my respectful opinion, it cannot be, unless the expert’s experience and training (i.e. professional education and feedback) have provided it with the requisite opportunities to also become constitutional experts in a given policy domain. This may potentially be present at more trial-like tribunals adjudicating questions of law on a daily basis (e.g. landlord and tenant boards), but such tribunals only represent a small minority of the Canadian administrative apparatus.

Behavioural research may even be interpreted to suggest that administrative bodies without legal expertise should ideally not be left to decide on protecting Charter rights at all until they have received special training. In Canada, administrators are permitted to undertake these Charter-related inquiries with the understanding that their decisions may be subject to judicial review. Such judicial review is said to promote, inter alia, a culture of justification and accountability. However, consider juxtaposing this with the following observation made by Lerner and Tetlock:

Predecisional accountability to an unknown audience [i.e. reviewing judges undertaking judicial review] will have no effect on bias if, even after increased attention to one’s decision process, no new ways of solving the problem come into awareness. Such is often the case when improvement on judgment task requires knowledge of formal decision rules … that are unfamiliar to the decision maker. … In essence, no amount of increased effort can compensate for lack of knowledge about how to solve problems that require special training [emphasis added].[109]

Here, again, it is worth mentioning the fictional telecom expert above, and reemphasizing the distinction between his knowledge of telecom statutes, technologies, and stakeholders on the one hand, and his knowledge of constitutional law as it should apply to the telecom industry on the other. Lerner and Tetlock would likely surmise that the two should not be conflated because the latter requires special training. Insofar as it does, then one potential implication is that experts in one administrative domain do not, in fact, increasingly improve their knowledge of constitutional law as it applies to that domain as they come across more cases implicating Charter rights. Such experience is not enough for them without being accompanied by special training. As Cameron pithily puts it, “true expertise means not only having information but understanding what to do with it”.[110] Unfortunately, the DorĂ© framework not seem to acknowledge or resolve this fully.

c. Some Recommendations Regarding Charter Rights in Administrative Law

The above would seem to suggest that a few recommendations are in order. Most generally, a more rigorous and probing (less deferential) standard of review should be applied to administrative decisions that implicate Charter rights––one that grants reviewing courts the latitude to reconsider the weight given to Charter rights in the original decision on the basis that the expert may not have had the expertise to allow it to grant the right its due weight. To support this, consider Cameron’s own position that, “where a tribunal has only non-expert insight––where it has … no special tools for analyzing … information, the judge will be just as able, if not better able, to assess the reasonableness of the tribunal’s inferences”.[111] Moreover, in response to Ponomarenko’s aforementioned concern regarding the presumptive weight given to the statutory objective within reasonableness review, the ideal standard of review for constitutional questions should also look more like a true Oakes proportionality test. Why? Firstly, as Macklin notes, “[a] proportionality analysis in the context of rights adjudication is not neutral as between rights … protected by the Charter and other interests, entitlements, or “values”. … a Charter right intrinsically ‘weighs’ more”.[112] If Charter rights truly do weigh more than other public interests, the standard of review should reflect this, and Charter rights should not, then, simply be thrown into a cauldron by the administrator to be stirred alongside other statutory objectives. Secondly, a full proportionality test would put the onus on the administrator to explain its position and would only protect its position and the interests it pursues insofar as it can be “demonstrably justified”.


Finally, the behavioural lens will be turned to one last aspect of Canadian administrative law: the duty placed on administrative bodies to provide full reasons for their decisions, and potential weaknesses surrounding the state of the law here. As Stratas JA has explained,  

[T]he more an administrative decision maker explains its decision and invoked expertise and specialized understandings in explicit reasons, the more the reviewing court is likely to find the administrative decision maker acted within its margin of appreciation. … In short, good reasons can be an admission ticket to deference.[113]

On its face, this seems to impose a standard whereby comprehensive reasons tackling every issue and factor at play in an administrative decision are provided by a body to the parties involved. However, as will be seen, the current law may permit less-than-comprehensive reasons, tackling less than all of the issues and factors at play in a decision, to be afforded deference under reasonableness review. From a behavioural standpoint, does this lead to problems?

a. Nurses’ Union and Deference to “Implicit” Reasons: A Problem?

The potential issue stems from the SCC’s discussion about tribunal reason-giving in Nurses’ Union.[114] Here, the Court reiterated its endorsement of Professor Dyzenhaus’ position that reviewing courts should be paying “respectful attention to the reasons offered or which could be offered in support of a [decision-maker’s] decision[115] and that a reviewing court applying reasonableness should “seek to supplement them [the reasons] before [seeking] to subvert them” [emphasis added].[116] A day before the Nurses’ Union decision was published, the SCC adopted a similar approach in ATA,[117] wherein Rothstein J stated “[i]f there exists a reasonable basis upon which the decision-maker could have decided as it did, the court must not interfere.”[118] Macklin believes that these cases seem to suggest that the Court is currently willing to retrofit the outcome reached by the decision-maker with judicially created reasons.[119] Likewise, Wildeman concludes that the current law provides that “it is not simply reasons and conclusions that are to be read together … but reasons, conclusions, and other contextual information available from the record (and perhaps other contextual sources)”.[120] In short, full reasons are not necessarily always required where the Court is willing and able to connect the dots on behalf of the administrator. Putting concerns aside which relate to how this approach may undermine administrative autonomy––and, in turn, legislative supremacy––does the behavioural perspective reveal problems with the reviewing courts’ willingness to do some of the decision-maker’s work for it? If an administrative body is not writing reasoned justifications comprising all relevant issues and factors to accompany its decisions with the understanding that reviewing courts will simply “supplement” incomplete reasons, does it have negative implications on the rigour of the decision-making itself?

b. Robust Reasons: A Means to an End and an End in Themselves?

It seems that both of the above questions can be answered in the affirmative. This conclusion can and has been drawn even without a strictly behavioural analysis. For instance, Daly has said that the position in NursesUnion regarding reasons “is inimical to good administration … because a decision-maker might be tempted to take a less rigorous approach to decision-making if it knows it can supplement its reasons later on”.[121] Speaking more generally about writing reasons, Sharpe JA also believes that “[t]he obligation to give reasons is much more than a legal formality.”[122]

There are admittedly many reasons as to why this is the case. For the purposes of this discussion, though, the central idea to consider is that the writing of robust reasons seems to have an impact on the decision-making mind. How so? For one, the more a decision-maker is forced to interact with and make inferences from an evidentiary record, the less opportunity there may be for overconfident reliance on expert intuition. The effect of a confirmation bias may also be reduced if the decision-maker is constrained to justify its decision in light of all facts and factors, rather than those salient ones that accord most well with its own interests. Speaking generally, Choi and Pritchard note how the prospect of judicial review of a decision should “discourage regulators from relying on heuristics because they must justify their decisions in writing to nonexperts”.[123] This may be true, but the discouraging effect would likely weaken if the decision-maker is also aware that (i) the reviewing court will assist it in supplementing existing reasons with others that could have been given in support of the decision, and that, as a result (ii) the decision-maker need not necessarily have its reasons touch upon all relevant facts or factors at play. In other words, the problem here is that Nurses’ Union may allow for decision-makers to produce incomplete reasons grounded on only parts of the evidentiary record. As soon as this is permitted, the spectre of the confirmation bias and reasoning based on overly confident intuitions may potentially reappear because the decision-maker is able to underemphasize or shun portions of the record that present weaknesses. To assess this more comprehensively, an empirical analysis surrounding reason-writing in the Canadian administrative context is well warranted.


a. More Robust Judicial Review

With the above decision-making pitfalls in mind, it becomes clear that, while the administrative state is undoubtedly composed of many much-needed experts, that expertise cannot always be relied upon in the way that a reasonableness standard of review based upon deference to expertise would seem to suggest. It is my intention here, then, to employ the behavioural research in order to specifically outline why a more rigorous reasonableness review may serve as a useful debiasing mechanism. Such a discussion should also be taken as a brief collection of ideas that should ideally be informing the thinking of the SCC as it grapples with the decision-writing for the current administrative law “trilogy” before it at the time of writing.

Both Eric Ip and Choi and Pritchard have separately argued that common law judicial review can exhibit many behaviourally sensible characteristics conducive to debiasing.[124] In particular, Ip explains that judicial review is most effective when it is “obliging decision makers to take an outsider perspective, to contemplate the opposite outcome to that which they prefer, to interact in the decision-making process with people of diverse backgrounds and biases, and to publicly defend their conclusions”.[125] Quoting Rachlinski, Choi and Pritchard comment, further, that biases are easier to spot in others than in oneself.[126] The problem in Canada, though, is that it cannot be said that a deferential standard which presumes adequate expertise is obliging decision-makers to do much of the exercises above to their full extent. This is not to say that they do not work long and hard to make decisions, as they most certainly do so. Nevertheless, the seemingly low level of judicial scrutiny and probing that is practiced by reviewing courts applying reasonableness––a standard which, by its very nature, does not require the administrator to make the correct decision––perhaps just leaves too much latitude for one-dimensional decision-making lacking a full appreciation for other important interests, rights, and nuances at play in a given case.

As alluded to above, with knowledge that a court will engage in a comprehensively probing analysis and be more reluctant in deferring to the decisions or supplementing the reasons of administrators, decision-making may improve. As Lerner and Tetlock observe in their study:

Simple conformity is not an option … when the views of the audience are completely unknown. Under such conditions, people … often engage in preemptive self-criticism … that is, they think in more self-critical, integratively complex ways in which they consider multiple perspectives on the issue and try to anticipate the objections that reasonable others might raise to positions that they might take.[127]

It is my belief that reviewing judges undertaking judicial review are akin to an audience whose views are unknown, in that their ultimate position regarding how the facts and the law in a given case will be interpreted is not, at all, known at the points during which the administrator itself is processing a decision. Insofar as this is true, then Lerner and Tetlock’s study confirms that decision-making should improve to the extent that an existing bias had initially resulted from a lack of effort, a lack of self-critical awareness of one’s judgment process, or both.[128] Without extrapolating too much, it seems as if a less deferential standard of review would, then, improve administrative efforts to grapple with the evidence without relying on mere intuition, and be more objective self-critical while doing so.

With that said, the above should not be taken to mean that judges themselves are not biased, as quite the contrary is the case. Research on decision-making biases and heuristics suggests that cognitive illusions plague many professionals,[129] and that judges are similarly plagued.[130] This should not detract from the fundamental role that judges play in undertaking review of administrative decisions, nor should it undermine the legitimacy of judicial review itself. To be rid of biases is fundamentally inhuman, and so the objective should be to recognize and manage human biases so that their effects may be minimized. Judicial review can help accomplish this, notwithstanding the individual pitfalls of judges themselves.

Furthermore, it would be ideal, in the context of the SCC currently contemplating a reconsideration or recalibration of the standard of review, for there to be an acknowledgment, at common law, that administrative expertise is not as reliable or sacrosanct as the post-CUPE and subsequent post–Dunsmuir jurisprudence has suggested. Put more pointedly, it would be ideal if a revised standard of review were to acknowledge what behavioural research has been demonstrating since the 1980s; that is, that decision-makers––experts included––make choices in a boundedly rational manner often susceptible to cognitive biases and heuristics that lead to unintended and initially hard-to-detect errors,  and that such problems are exacerbated when the very scope of an administrator’s expertise is jurisprudentially  stretched or expanded to cover areas of decision-making within which an expert has not, well, proven to be an expert. In other words, expertise is not institutionally inherent, absolute, or unwaveringly reliant; rather, it is developed and honed at the individual level, it is relative, and it is sometimes inextant.

Overall, then, I view my proposal here as one advocating for the SCC to consider a more “principled approach”––this term specifically being used here with the intention that its meaning be in line with how the SCC has used it in its evidence law jurisprudence over the last few decades[131]––wherein a foundational assumption or presumption about an administrator is to be given effect upon judicial review of that administrator’s decision only insofar as that principle functionally and substantively holds within the circumstances. As such, if it is the case that the assumed or presumed presence of administrative expertise is inappropriate, untenable, or has otherwise been rebutted in a given case, then curial deference on the basis of such expertise should concomitantly fall away. Of course, this stance then necessitates several other accompaniments. For one, it would mean that the SCC should no longer be able to assert that expertise inheres in a tribunal as an institution and, as a result, relative expertise of specific administrators in specific circumstances will have to be assessed in some reliable manner. This, in turn, would mean that a much more contextual and administrative-body-, administrator-, and/or issue-dependent analysis would likely have to be undertaken by courts upon judicial review. It is beyond the scope of this paper to discuss the full extent of the costs and benefits that a more contextual and probing standard of review would trigger. Yet, insofar as such a standard of review would become more so principled and more so adherent to the empirical bases for relying on expertise, then the change would be most welcome from my current vista.

Of course, implementing proposals such as the one advocated above by means of a trilogy-induced change to the common law at the SCC may be much more difficult than I have hitherto acknowledged. Indeed, this is, in part, why Daly seems so pessimistic about the propensity for the current administrative law trilogy to produce fundamental change or renewal[132] (see Part II, above). Consider briefly some barriers to fundamental changes that several notable commentators have mentioned in writings about the current trilogy before the SCC which are particularly relevant in light of the discussion in this paper:

  • Factionalism. There remain well-entrenched factions on the Court stemming from “deep-seated disagreement about the foundations of judicial review of administrative action”.[133] As Daly puts it, “without agreement about foundational issues––the basis on which doctrine can be built––there will be no lasting agreement about day-to-day issues with the application of the Court’s doctrine”.[134] In other words, a fundamental change to the current standard of review is unlikely because there are simply too many doctrinal divisions on the Court that cannot be assuaged effectively enough in the context of nine judges coming together to wrap their minds around writing the decisions in this trilogy. Partly due to such disagreements, Macklin, almost immediately after making her own oral submissions to the SCC in this trilogy in December 2018, opined that “I do not think the Court will do anything in these appeals that will affect, in any substantive way, the standard of review”.[135]
  • The Spectre of Merely Piecemeal Changes. Related to the above is the notion that “[t]he day-to-day problems which prompted the general frustration … are not problems that can be resolved by performing a nip and tuck to the Dunsmuir framework”.[136] The problem that follows from this, though, is that, if foundational and fundamental adjustments to the standard of review are required, but such adjustments are not possible given an SCC home to great factionalism (per the above), then the trilogy may ultimately only end up producing a “nip and tuck” of Dunsmuir.  This is not enough, especially given the arguments made throughout (and beyond) this paper that what is needed is a reassessment and reconfiguration of the very foundations and factors that undergird current jurisprudence surrounding deference and expertise, inter alia. Once again, I reference, as a support, Macklin’s experiences while both making and observing the oral submissions to the Court in this trilogy which encouraged the justices to consider a doctrinal overhaul of the standard of review,  during which, as she puts it, “I grew increasingly discouraged by the interaction between the Court and the procession of the parties and interveners who came before it. I am now less optimistic about what this Court … can do with the hot mess that standard of review has become”. [137] Later on in this same piece, she falls even more deeply into defeatism: “[s]omething has to give, and I worry that the Court may be flattered into believing that tinkering will do. It won’t.[138]
  • Institutional Limitations. There are potentially debilitating limitations surrounding the SCC’s institutional competence and capacity to undertake a reappraisal of the standard of review that is comprehensive enough. As Daly notes, “[r]esolving the wide range of issues with all appropriate information to hand may, frankly, be the task of a Royal Commission.”[139] In other words, the Court may not be able to formulate a standard of review that is comprehensively informed by contemporary policy considerations, sound theory, and empirically-informed foundational principles because it simply does not have enough time and resources to do so. The implication here is that such an undertaking should be one that legislatures should consider exploring. This is, perhaps, the most problematic limitation for the purposes of this paper. How so? For one, producing behaviourally-informed empirical analyses of the effectiveness, scope, and overall sacrosanctity of administrative decision-making expertise is something that should require significant time and specialized knowledge from the likes of professionals and experts from realms outside of the legal community. It should ideally also involve the use of studies which take time, require money, and involve consultation and partnership with social scientists without an interest in the ultimate result in these three appeals. Moreover, interpreting and effectively implementing the results of such analyses is something that nine judicial minds trained in the art of legal reasoning may not be able to do as effectively as a legislature armed with, for instance, many more investigative, committee, advisory-seeking, and public-comment-related powers. Yet, in the institutional context in which the Court finds itself, “[e]ven three days, assisted by the country’s ablest administrative lawyers, are unlikely to be enough.”[140]

And so, while, with respect to formulating a more robust judicial review, we may currently find ourselves at a crucial juncture that may very well provide opportunity for change, the above issues may ultimately turn this juncture into yet another impasse. If this ends up being the case, then my position would be that, rather than wait another decade for the standard of review to be reconsidered by the SCC, legislatures should instead strongly consider ways in which they might be able to assist courts, lawyers, and administrators in exploring alternative avenues of doctrinal and systemic change.

b. Algorithms and Artificial Intelligence

Statistical inquiries, algorithmic analyses, and other computational supports that are calibrated to calculate the correct factors in a given case or study may be able to do so very efficiently and relatively objectively. For experts, it would mean that there is less opportunity for reliance on unhelpful intuition,[141] and this can be a good thing. Coglianese and Lehr have gone so far as to posit that “human decisionmaking is highly susceptible to bias and error, so algorithmic adjudication could actually enhance governmental fairness, not detract from it”.[142] While reliance on technology to adjudicate a decision in full may be overly controversial[143] and present its own problems, the use of technology for consultative or predictive support would likely be a most welcome development for all stakeholders involved in an administrative decision. Indeed, there have been studies comparing expert predictions to statistical or algorithmic ones.[144] In most cases, “the accuracy of experts was matched or exceeded by a simple algorithm”.[145] Government agencies in the US, such as the EPA and the SEC, are already using machine learning to assist them in decision-making.[146]

c. Internal Review Mechanisms

Beyond technological supports, there are other internal review mechanisms that may be worthwhile. Many administrative bodies have semi-independent groups or departments of analysts, lawyers, and other professionals whose job it is to provide advice throughout the decision-making process. Permitting these groups to participate as often and as candidly as possible is something that should be promoted across the administrative state. Furthermore and related to this, internal adversarial review should also be considered.[147] For instance, mock trials in which team members are constrained to consider alternative interpretations and conclusions, if taken seriously, may help reveal logical leaps, mistaken assumptions, underemphasized information, or overemphasized anecdotes.

d. External Review Mechanisms

External consultation, audits, and reviews should also be considered as much as is practicable. As Sharpe JA notes how “[s]pecialists have a tendency to become too familiar and too comfortable with the way things have always been done. An outside perspective can bring a breath of fresh air.”[148] Indeed, not only has it generally been shown that  the quality of the evidence review may be improved by external expert input,[149] there is also some evidence that when external experts make decisions on behalf of others rather than for their own organization, this mitigates their own cognitive biases.[150] As Zamir and Teichman remark, “people are presumably more dispassionate when they make decisions for others than for themselves”.[151] The authors describe that this may be because there is “greater psychological distance” present when someone is called upon to help others decide rather than himself, and that “the greater the psychological distance, the more abstract people’s thinking tends to be”.[152] In other words, not only would external review mechanisms potentially allow for more ideas and perspectives to be considered, but the people providing these ideas and perspective may be doing so from a more objective outlook.


Curial deference in the context of the reasonableness standard of review––the standard of review that presumably applies to an administration decision––should not be resting on certain assumptions about expertise if they are found to be, at least partially, untenable. The above discussion, in which I have sought to employ the wisdom of behavioural economics to the current emanation of reasonableness review, has demonstrated that some of its central underlying assumptions regarding the quality and the scope of administrative expertise are indeed untenable. Experts are prone to engage in biased decision-making that is less than comprehensive in many ways. The administrative decisions that come from these experts may, therefore, be more so unreasonable than previously anticipated. Given the pervasiveness of the administrative state, this simply does not bode well for Canadians.

As such, empirical studies testing the positions I have arrived at in this paper are well warranted. Government research into methods of administrative decision-making that minimize the influence of cognitive biases and heuristics would also be most welcome. Other countries, such as the UK, have begun to fund government organizations to explore these very issues.[153] Canada would do well to consider doing the same. Furthermore, as has been briefly discussed above, insofar as such empirical studies reveal more limits surrounding reliance on administrative expertise, then concomitant changes should, at some point, be made to the common law or, even, to  the Canadian statutory law governing administrative rules and principles to be applied by courts upon judicial review of executive action. The reasonableness standard of review should be a principled one, attuned to the actual administrative circumstances on the ground in a given case. If the primary basis for curial deference is the presence of administrative expertise, then a court should not be deferring to the administrator when such expertise is not extant or not relatively strong or reliable. As Macklin has pointedly put it: “many administrative decision makers are indeed experts at what they do. But deeming all administrative decision makers as experts will not make the remainder into experts.”[154] The SCC currently has the opportunity to consider acknowledging such nuances and limitations as it continues to work toward the completion of its reasons for this most recent administrative law trilogy of cases. It is thus with great curiosity and hope that many of us await what the Court has to say.


[1] Canadian Broadcasting Corp v SODRAC 2003 Inc, 2015 SCC 57 at para 185, [2015] 3 SCR 615, Abella J, dissenting [SODRAC].

[2] Eyal Zamir & Doron Teichman, Behavioral Law and Economics (Oxford, UK: Oxford University Press, 2018) at 394 [Zamir & Teichman].

[3] Eric C Ip, “Debiasing Regulators: The Behavioural Economics of US Administrative Law” (2017) 46:3 Comm L World Rev 171 at 171 [Ip].

[4] Ibid at 172.

[5] Edmonton (City) v Edmonton East (Capilano) Shopping Centres Ltd, 2016 SCC 47 at para 33, [2016] 2 SCR 293 [Capilano]; see also Hillary Evans Cameron, “Substantial Deference and Tribunal Expertise Post-Dunsmuir: A New Approach to Reasonableness Review” (2014) 27:1 Can J Admin L & Prac 1 at 7–8 [Cameron].

[6] Dunsmuir v New Brunswick, 2008 SCC 9 at para 47, [2008] 1 SCR 190 [Dunsmuir].

[7] Bell Canada, et al v Attorney General of Canada (2019), SCC docket no 37896, 2018 CarswellNat 2186, 2187 [Bell]; National Football League, et al v Attorney General of Canada (2019), SCC docket no 37897, 2018 CarswellNat 2188, 2189 [NFL]; Minister of Citizenship and Immigration v Alexander Vavilov (2019), SCC docket no 37748, 2018 CarswellNat 2127, 2128 [Vavilov].

[8] Namely, in ibid, Bell and Vavilov.

[9] Ibid.

[10] Paul Daly, “The Supreme Court of Canada’s Administrative Law Trilogy: What to Expect?” [blog post] Administrative Law Matters (November 26, 2018), online: [Daly Blog 1].

[11] Ibid.

[12] Audrey Macklin, “Standard of Review: Back to the Future?”, in Colleen Flood & Lorne Sossin, eds, Administrative Law in Context, 3rd ed (Toronto: Emond Montgomery, 2018) at 409 [Macklin, “Back to the Future?”]; see also Lauren Wihak, “Wither the Correctness Standard of Review? Dunsmuir, Six Years Later” (2014) 27:2 Can J Admin L & Prac 173 at 174–175 [Wihak].

[13] Dunsmuir, supra note 6 at para 49.

[14] Gus Van Harten et al., Administrative Law: Cases, Text, and Materials (Toronto: Emond Montgomery, 2015) at 674 [Van Harten].

[15] Dr Q  v College of Physicians and Surgeons of British Columbia, 2003 SCC 19 at para 29, [2003] 1 SCR 226 (BC).

[16] CUPE, Local 963 v New Brunswick Liquor Corp (1979), [1979] 2 SCR 227, 97 DLR (3d) 417. 

[17] Van Harten, supra note 14 at 673.

[18] Pushpanathan v Canada (Minister of Employment & Immigration) (1998), [1998] 1 SCR 982, 160 DLR (4th) 193.

[19] Canada (Director of Investigation & Research) v Southam Inc (1997), [1997] 1 SCR 748, 144 DLR (4th) 1.

[20] Macklin, “Back to the Future?”, supra note 12 at 397.

[21] Cameron, supra note 5 at 5. see also Lorne Sossin, “Empty Ritual, Mechanical Exercise or the Discipline of Deference? Revisiting the Standard of Review in Administrative Law” (2003) 27 Advocates’ Q 478 at 490–492.

[22] Dunsmuir, supra note 6 at para 54.

[23] Donald J Jordan, “Deference to Administrative Tribunal Interpretation of their Home Statute: An Obvious Observation (2017) 30:3 Can J Admin L & Prac 389 at 392 [Jordan].

[24] Macklin, “Back to the Future?”, supra note 12 at 418. Interestingly, as Macklin observes: “it is unclear whether demonstrated lack of expertise is capable of rebutting a presumption of reasonableness, or useful in calibrating how reasonableness should be applied” (418).

[25] Capilano, supra note 5 at para 33; see also Cameron, supra note 5 at 7–8.

[26] Sheila Wildeman, “Making Sense of Reasonableness”, in Colleen Flood & Lorne Sossin, eds, Administrative Law in Context, 3rd ed (Toronto: Emond Montgomery, 2018) at 464 [Wildeman].

[27] Zamir & Teichman, supra note 2 at 141.

[28] Ibid at 142.

[29] Cass Sunstein, “Behavioral Law and Economics: A Progress Report” (1999) 1:1/2 Am L & Econ Rev 115 at 115 [Sunstein, “A Progress Report”].

[30] Cass Sunstein, “The Storrs Lectures: Behavioral Economics and Paternalism” (2013) 122:7 Yale LJ 1836 at 1829 [Sunstein, “Paternalism”]. Sunstein served in this role from 2009 to 2012.

[31] Ip, supra note 3 at 172.

[32] Jonathan Baron, “Heuristics and Biases”, in Eyal Zamir and Doron Teichman, The Oxford Handbook of Behavioral Economics and the Law (Oxford, UK: Oxford UP, 2014) 1 at 4.

[33] Daniel Kahneman, Thinking, Fast and Slow (Toronto: Anchor Canada, 2011) [Kahneman].

[34] Ibid at 20–21.

[35] Ibid at 21; see also Sunstein, “Paternalism”, supra note 30.

[36] Ibid at 24; see also Sunstein, “Paternalism”, supra note 30.

[37] Sunstein, “A Progress Report”, supra note 29 at 121.

[38] Ibid at 123.

[39] Kahneman, supra note 33 at 140; See generally Colin Kuehnhanss et al., “Choice in Politics: Equivalency Framing in Economic Policy Decisions and the Influence of Expertise” (2015) 40 Eur J L & Econ 360.

[40] Sunstein, “A Progress Report”, supra note 29 at 146.

[41] Zamir & Teichman, supra note 2 at 515.

[42] Sunstein, “A Progress Report”, supra note 29 at 164.

[43] Van Harten, supra note 14 at 675. See, for example, Rothsten J’s statement in Alberta (Information and Privacy Commissioner) v Alberta Teacher’s Association, 2011 SCC 61 at para 1, [2011] 3 SCR 654 (Alta) that, in creating administrative tribunals, “legislatures confer decision-making authority on certain matters to decision-makers who are assumed to have specialized expertise with the assigned subject matter”.

[44] Van Harten, supra note 14 at 674.

[45] Jordan, supra note 23 at 393

[46] Jeffrey Rachlinski & Cynthia Farina, “Cognitive Psychology and Optimal Government Design” (2002) 87 Cornell L Rev 549 at 579 [Farina].

[47] Robert J Sharpe, Good Judgment: Making Judicial Decisions (Toronto: University of Toronto Press, 2018) at 27 [Sharpe].

[48] Ibid at 36.

[49] Farina, supra note 46 at 558.

[50] James Cooper & William Kovacic, “Behavioral Economics: Implications for Regulatory Behavior” (2012) 41 L Regul Econ 41 at 44 [Cooper].

[51] Farina, supra note 46 at 558.

[52] Ibid at 558.

[53] Cooper, supra note 50 at 49.

[54] Ibid at 50.

[55] Kahneman, supra note 33 at 240.

[56] Ibid at 241.

[57] Cooper, supra note 50 at 50.

[58] Zamir & Teichman, supra note 2 at 398.

[59] Farina, supra note 46 at 559.

[60] Constitution Act, 1867, RSC 1985, App II, No 5, s 96.

[61] Wildeman, supra note 26 at 468–471.

[62] NB: The discussion below proceeds under the assumption that the reader is already acquainted with the noted behavioural concepts. For general information on the details of each of the biases and heuristics below and others, see generally Zamir & Teichman, supra note 2.

[63] Ip, supra note 3 at 175.

[64] Cooper, supra note 50 at 46.

[65] Ip, supra note 3 at 175.

[66] Ip, supra note 3 at 175.

[67] See Stephen Choi & AC Pritchard, “Behavioral Economics and the SEC” (2003) 56:1 Stanford L Rev 1 at 30 [Choi], where Choi and Pritchard discuss the potential prevalence of this bias and others within the US Securities and Exchange Commission (SEC).

[68] Ip, supra note 3 at 175.

[69] Ibid at 175–176.

[70] Ibid at 176.

[71] Cooper, supra note 50 at 47.

[72] For a discussion and an empirical analysis of this in the context of lawyers, see Jeffrey Rachlinski & Andrew Wistrich, “How Lawyers’ Intuitions Prolong Litigation” (2013) 86 S Cal L Rev 101 at 124­­­­–133.

[73] Choi, supra note 67 at 28.

[74] Ip, supra note 3 at 176.

[75] Farina, supra note 46 at 579.

[76] Ip, supra note 3 at 176.

[77] Kahneman, supra note 33 at 60.

[78] Ibid at 240.

[79] Ip, supra note 3 at 185.

[80] Ibid at 176.

[81] Baker v Canada (Minister of Citizenship and Immigration) (1999), [1999] 2 SCR 817.

[82] Ibid at para 5.

[83] Choi, supra note 67 at 33­­–34.

[84] Ip, supra note 3 at 176; see also ibid at 33­­–34.

[85] Ibid at 176; see also Michael Hallsworth et al., “Behavioural Government: Using Behavioural Science to Improve how Governments Make Decisions”, BIT report (London, UK: Behavioural Insights Team, 2018), at 9 [BIT Report].

[86] Zamir & Teichman, supra note 2 at 122.

[87] Canadian Charter of Rights and Freedoms, s 1, Part I of the Constitution Act, 1982, being Schedule B to the Canada Act, 1982 (UK), 1982, c 11 [Charter].

[88] Dunsmuir, supra note 6 at para 58.

[89] Ibid.

[90] See the majority and concurring judgments in Multani v Commission scolaire Marguerite- Bourgeoys, 2006 SCC 6, [2006] 1 SCR 256.

[91] Macklin, “Back to the Future?”, supra note 12 at 423.

[92] Doré v Barreau du Quebec, 2012 SCC 12, [2012] 1 SCR 395 [Doré].

[93] Ibid at para 29.

[94] Ibid at para 7; see also Audrey Macklin, “Charter Right or Charter-Lite?: Administrative Discretion and the Charter” (2014) 67 SCLR 561 at 570 [Macklin, “Charter-Lite?”].

[95] Doré, supra note 92 at para 36.

[96] Ibid at para 54.

[97] Ibid at para 48.

[98] Ibid at para 56. A discussion of the constitutional issues here is beyond the scope of this paper.

[99] Trinity Western University v The Law Society of British Columbia, 2016 BCCA 423 at para 82, 2017 CanLII 8574.

[100] Wildeman, supra note 26 at 496.

[101] Iryna Ponomarenko, “Tipping the Scales in the Reasonableness-Proportionality Debate in Canadian Administrative Law” (2016) 21 Appeal Rev Current L & L 125 at 134 [Ponomarenko].

[102] Ibid at 134.

[103] Ip, supra note 3 at 175–176.

[104] Cooper, supra note 50 at 41.

[105] Macklin, “Charter-Lite?”, supra note 94 at 576.

[106] Kahneman, supra note 33 at 241.

[107] Ibid at 242.

[108] Doré, supra note 92 at para 54.

[109] Jennifer Lerner & Philip Tetlock, “Accounting for the Effects of Accountability” (1999) 125:2 Psychological Bulletin 255 at 263 [Lerner].

[110] Cameron, supra note 5 at 15.

[111] Ibid at 16.

[112] Macklin, “Charter-Lite?”, supra note 94 at 573.

[113] David Stratas, “The Canadian Law of Judicial Review: A Plea for Doctrinal Coherence and Consistency” (2016) 42:1 Queen’s LJ 27 at 52 [Stratas].

[114] Newfoundland and Labrador Nurses’ Union v Newfoundland and Labrador (Treasury Board), 2011 SCC 62, [2011] 3 SCR 708 [Nurses’ Union].

[115] Ibid at para 12.

[116] Ibid at para 12; see also Cameron, supra note 5 at 9–10.

[117] Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association, 2011 SCC 61, [2011] 3 SCR 654 [ATA].

[118] Ibid at para 53.

[119] Macklin, “Back to the Future?”, supra note 12 at 428.

[120] Wildeman, supra note 26 at 470.

[121] Paul Daly, “Reasonableness and Reasons in Administrative Law: Delta Air Lines Inc v Lukacs,” (2018) 31:2 Can J Admin L & Prac 210 at 214.

[122] Sharpe, supra note 47 at 134.

[123] Choi, supra note 67 at 38.

[124] Ip, supra note 3 at 178; Choi, supra note 67 at 37.

[125] Ibid at 178.

[126] Choi, supra note 67 at 37.

[127] Lerner, supra note 109 at 257.

[128] Ibid at 263.

[129] Chris Guthrie, Jeffrey Rachlinski & Andrew Wistrich, “Judging by Heuristic: Cognitive Illusions in Judicial Decision Making,” (2002) 86 Judicature 44 at 44.

[130] Ibid at 45.

[131] See generally R v Khan, [1990] 2 SCR 531, 59 CCC (3d) 92; R v B(KG), [1993] 1 SCR 740, 79 CCC (3d) 257; R v Khelawon, [2006] 2 SCR 787, 215 CCC (3d) 161.

[132] Daly Blog 1, supra note 10.

[133] Ibid.

[134] Ibid.

[135] Mark Mancini, “Lowering Expectations: The Supreme Court’s Standard of Review Cases” [blog post] Double Aspect (November 29, 2018) online: [Mancini Blog 1].

[136] Daly Blog 1, supra note 10.

[137] Audrey Macklin [guest post], “Audrey Macklin on the Supreme Court of Canada’s Administrative Law ‘Trilogy’” [blog post] Administrative Law Matters (December 14, 2018) online: [Macklin Blog]; see also Mark Mancini, “The Administrative Law ‘Trilogy’: The Stare Decisis Trap” [blog post] Double Aspect (December 7, 2018) online: [Mancini Blog 2]; and Mark Mancini, “The New Administrative Law” [blog post] Double Aspect (April 30, 2019) online: [Mancini Blog 3].

[138] Ibid, Macklin Blog.

[139] Daly Blog 1, supra note 10.

[140] Paul Daly, “A Tale of Two Courts: Administrative Law in the SCC and the UKSC”[blog post] Administrative Law Matters (December 20, 2018) online: [Daly Blog 2].

[141] Zamir & Teichman, supra note 2 at 128.

[142] Cary Coglianese & David Lehr, “Improving the Administrative State with Machine Learning” (2017) 42:4 Admin & Reg L News 7 at 7 [Coglianese].

[143] Kahneman, supra note 33 at 229.

[144] Ibid at 222.

[145] Ibid at 223.

[146] Coglianese, supra note 142 at 8.

[147] Cooper, supra note 50 at 53.

[148] Sharpe, supra note 47 at 39.

[149] BIT Report, supra note 85 at 11.

[150] Zamir & Teichman, supra note 2 at 117.

[151] Ibid at 119.

[152] Ibid.

[153] See generally BIT Report, supra note 85. The Behavioural Insights Team was co-founded by the UK Government in 2010 along with several nongovernmental organizations.

[154] Macklin Blog, supra note 137; see also Mark Mancini, “The Dark Art of Deference” [blog post] Double Aspect (March 6, 2018) online: [Mancini Blog 4].