Decision considered: Alberta (Information and Privacy Commissioner) v Alberta Teachers Association, 2011 SCC 61
In its recent decision reversing the Alberta Court of Appeal’s decision in Alberta (Information and Privacy Commissioner) v Alberta Teachers’ Association (Teachers’ Association), 2010 ABCA 26, the Supreme Court of Canada made significant statements with respect to issues of administrative law. In particular, a majority of the Court held:
- When an issue is not raised before an administrative decision-maker it may nonetheless be raised in an application for judicial review. A court may, however, exercise its “discretion not to consider an issue raised for the first time on judicial review where it would be inappropriate to do so” (para 22).
- In such cases deference may still be granted: “Where the reviewing court finds that the tribunal has made an implicit decision on a critical issue, the deference due to the tribunal does not disappear because the issue was not raised before the tribunal” (para 50).
- In order to be deferential in such circumstances, the court may take into account the reasons that the administrator could have given had the issue been put before it. If a “reasonable basis for the decision is apparent to the reviewing court” then that will suffice (para 55). The court may also look at reasons offered by the administrative decision-maker on the issue in other cases to determine whether the decision-maker’s approach to the issue is reasonable. In some circumstances the court may remit the matter to the decision-maker to allow reasons to be prepared.
- Finally, and most significantly, a majority of the Court, in reasons prepared by Justice Rothstein, called into question the ability to identify a “true question of jurisdiction” to which deference should not be granted. Justice Rothstein stated that he was “unable to provide a definition of what might constitute a true question of jurisdiction” (para 42).
- Justice Rothstein held that when an administrative decision-maker interprets its home statute it is presumptively entitled to deference (para 34). Deference will not be offered where the interpretation raises constitutional questions, a question regarding the jurisdictional lines between tribunals or a question of law “that is of central importance to the legal system as a whole and that is outside the adjudicator’s expertise” (para 43). If a party claims that deference is not owed because the matter is a true question of jurisdiction, that party will “be required to demonstrate why the court should not review a tribunal’s interpretation of its home statute on the deferential standard of reasonableness” (para 39).
- Finally, Justice Rothstein held that once a deferential standard has been identified, it is not necessary to question further how deferential the court should be: “Once it is determined that a review is to be conducted on a reasonableness standard, there is no second assessment of how intensely the review is to be conducted” (para 47).
The last three aspects of the Justice Rothstein’s judgment are the most significant, and they arediscussed in this blog. Justice Binnie (with Justice Deschamps) and Justice Cromwell each wrote concurring judgments taking exception to Justice Rothstein’s approach to jurisdictional questions. Further, Justice Binnie again expressed his disagreement with the view that it was possible to take a singular approach to deferential review, suggesting that the extent of a court’s deference must necessarily vary with the nature of the decision being considered.
After summarizing the background and outcome of the case (with which all the judges agreed), I analyze in greater detail the dispute between the judges as to the correct approach to jurisdictional questions, and as to how deference ought to be conducted given the divergent nature of the decisions courts are asked to review.
In my view, each of the judges recognizes the difficulty with an ex ante identification of a true question of jurisdiction, and none suggests a definition of such questions that permits them to be identified acontextually. The difference between the judgments is thus not, at its heart, about whether true questions of jurisdiction exist. Rather, the difference between them is as to the extent to which courts should defer to legal decisions made by administrative decision-makers. The difference is one of principle, going to the heart of administrative law’s quandary: who should decide? I will suggest that that disagreement cannot be resolved, because both positions are justified in principle. A legal system can properly respect rule of law and legislative supremacy either when courts largely defer to legal decisions made by administrative decision-makers, or when they do so less readily. What may be important, however, is that the Court figures out the approach it wants to take, and sticks with it. Constantly shifting and re-evaluating these questions – which the Court is now entering its 4th decade of doing – is unhelpful to the courts below, to administrative decision-makers, to lawyers preparing their cases and advising clients, and, most significantly, to the parties affected by those decisions.
The Teachers’ Association case arose out of complaints brought to the Office of the Information and Privacy Commissioner alleging violations of privacy by the Alberta Teachers’ Association (para 6). The Commissioner started an investigation, which pursuant to subsection 50(5) of the Personal Information Protect Act, SA 2003, c P-6.5, (“PIPA”), was required to complete within 90 days of the complaint being received. The investigation was not completed within 90 days. Further, the Commissioner did not exercise its statutory power to extend the 90 day time period until 22 months had passed (s 50(5)).
The issue raised on judicial review of the case was thus as follows: can the power to extend the time limit imposed by PIPA be exercised after the time limit has passed? The issue was not raised before the Commissioner. However, in a number of other cases the Commissioner and its delegated adjudicators had addressed the issue, both in the context of PIPA and in the context of the similar provision contained in the Freedom of Information and Protection of Privacy Act, RSA 2000, c F-25 (para 56). In those decisions the Commissioner had “provided a consistent analysis” (para 56) and had concluded that time extensions could be granted after the expiry of the 90 days period.
On judicial review, the Alberta Court of Queen’s Bench and the Alberta Court of Appeal (Berger J. dissenting) were of the view that the statutory limit could not be extended after the expiry of the 90 day period. The Court of Queen’s Bench viewed the question as jurisdictional, and applied a correctness standard of review (para. 11). The Court of Appeal decided the matter de novo on the grounds that the adjudicator had never considered it (para 15). Berger JA dissented on the basis that, properly interpreted, subsection 50(5) permitted the time period to be extended even after it had expired. He also “found that quashing the adjudicator’s order without the benefit of reasons compromised judicial review” (para 19).
The Supreme Court of Canada unanimously granted the appeal, holding that the Commissioner’s decision was entitled to deference, and that its implicit interpretation of subsection 50(5) of PIPA was reasonable. The points of unanimity in the Court’s decision are summarized above; my focus here will only be on the matters on which the judgments of the Court diverged – on the existence of true questions of jurisdiction, and the methodology for deferential review.
As noted, the Court of Queen’s Bench viewed the issue of time lines as a matter of jurisdiction, of “whether the Commissioner had lost jurisdiction over the inquiry as a result of his failure to complete the inquiry within the timelines setout in s. 50(5) of the PIPA” (para 11). None of the judges at the Supreme Court agreed; all thought the interpretation of s. 50(5) was a matter with respect to which the Commissioner was entitled to deference. The Court split, however, on the principles underlying that determination.
For a majority of six, Justice Rothstein called into question the usefulness of trying to determine whether an issue raised on judicial review can be categorized as a “true question of jurisdiction” (para 33). He noted the dangers to respect for ?? administrative expertise posed by expansive identification of questions as jurisdictional. He observed that, post Dunsmuir (Dunsmuir v New Brunswick, 2008 SCC 9), the Court had not identified “a single true question of jurisdiction” (para 33).
Rothstein J. suggested in particular that a court should be most reluctant to categorize matters as jurisdictional when they involve interpretation of a home statute. He noted that any such interpretation may arguably be characterized as jurisdictional, since it necessarily “involves the determination of whether it [the decision-maker] has the authority or jurisdiction to do what is being challenged” (para 34). Yet it is clear from the case law that the modern court does not normally view home statute interpretation or application as raising jurisdictional questions. Instead, the jurisprudence suggests that when a tribunal is interpreting its “own statute or statutes closely connected to its function, with which it will have particular familiarity,” a court should presume that the matter is one of statutory interpretation and, consequently, only subject to deferential review (para 34, quoting Dunsmuir). A party who would seek to rebut that presumption bears the burden of proof to establish “why the court should not review a tribunal’s interpretation of its home statute on the deferential standard of reasonableness” (para 39). Rothstein J. argued that this presumption of deference “was implicitly already established in Dunsmuir” (para 41).
Rothstein J. went so far as to suggest that “it may be that the time has come to reconsider whether, for purposes of judicial review, the category of true questions of jurisdiction exists and is necessary to identifying the appropriate standard of review” (para 34). He noted that he was unable to come up with any definition that could permit satisfactory identification of such questions. Absent such a definition, the concept of jurisdictional questions will simply result in confusion (para 42). Rothstein J. rejected the view that retaining the category of jurisdictional questions is important to maintain the rule of law. The standard of correctness would remain for any interpretation of a home statute which raises a “question of law that is of central importance to the legal system as a whole and that is outside the adjudicator’s expertise,” as well as for questions about the competing jurisdiction of different tribunals. He criticized Binnie J.’s judgment insofar as it would permit correctness review on questions of general importance to the legal system as a whole, whether or not they fall within the expertise of the decision-maker.
Rothstein J. also disagreed with Binnie J. about the idea of varying approaches to deferential review, stating that “Once it is determined that a review is to be conducted on a reasonableness standard, there is no second assessment of how intensely the review is to be conducted” (para 47). Review is contextual, but “there is no determination of the intensity of the review with some reviews closer to a correctness review and others not” (para 47).
Writing alone, Justice Cromwell disagreed with Rothstein J.’s position on jurisdiction. In his view jurisdictional questions exist, insofar as there are questions “with respect to which the courts are obliged to substitute their understanding of the correct answer for the tribunal’s understanding of the correct answer” (para 93). Those questions can arise when an administrative decision-maker is interpreting its home statute as well as otherwise; the point is simply this: who did the legislature intend to decide the question? Cromwell J. did not think it was necessary to come up with a definition of jurisdictional questions that could be used to identify such questions ex ante. Rather, legislative intent should be determined by the factors set out in the standard of review analysis, and the existence of a jurisdictional question – one requiring correctness review – established accordingly:
These are the concrete criteria, clearly established by the Court’s jurisprudence, which are used to identify questions that are reviewable for correctness because the legislature intended the courts to have the last work on what constitutes a ‘correct’ answer. These questions may be called jurisdictional (para 97)
Cromwell J. acknowledged that while “the language of jurisdiction” or “vires” might be unhelpful in the standard of review analysis,” that does not remove the need to inquire as to which standard of review applies (para 102). That a question relates to a home statute is not determinative.
Binnie J., writing for himself and Deschamps J., viewed himself as occupying a “middle ground” (para 78). In his view the “concept of jurisdiction is fundamental to judicial review” but the idea of a “true question of jurisdiction” does little practical work in deciding how deferential a Court should be (para 79 and 80). Further, Binnie J. saw neither Rothstein J.’s presumption, nor Cromwell J.’s search for legislative intent, as particularly useful for resolving the matter. In Binnie J.’s view, a decision-maker is entitled to deference when interpreting its home statute if the matter “is within its expertise and does not raise issues of general legal importance” (para 83). But if the matter is of general legal importance – even if not of “central importance to the legal system as a whole”- deference should not be given (para 84). “After all, some administrative decision makers have considerable legal expertise and resources. Others have little or none” (para 84).
Binnie J. further suggested that identifying whether a decision is reasonable “is context specific and varies with the circumstances including the nature of the issue under review” (para 85). In cases of legal interpretation, it may often be that deference is given, but that such deference is “not far removed from a correctness analysis” (para 85). By contrast, matters of “general policy or broad discretion” will be reviewed with a “much less aggressive attitude” (para 86).
In my view, these judgments have greater common ground than they themselves acknowledge. Most significantly, none of the judgments suggest a definition or categorization of jurisdictional questions that would allow them to be identified independently of other aspects used in determining standard of review. In that sense, I think, all of the judgments move away from Dunsmuir, which tried to define when matters could be considered to be truly jurisdictional:
“Jurisdiction” is intended in the narrow sense of whether or not the tribunal had the authority to make the inquiry. In other words, true jurisdiction questions arise where the tribunal must explicitly determine whether its statutory grant of power gives it the authority to decide a particular matter (Dunsmuir at para 59)
Here, by contrast, all of the judgments focus on determining when correctness review is required, without trying to categorize or independently identify some questions as jurisdictional. Rothstein J. would adopt a rebuttable presumption with respect to interpretation of the decision-maker’s home statute. Binnie J. would focus on whether a legal question is one of general importance to the legal system. Cromwell J. would use the standard of review analysis in cases of uncertainty to identify legislative intention. None of the judges suggest that one can look independently at the nature of the issue before the court and determine whether it requires that “the tribunal must explicitly determine whether its statutory grant of power gives it the authority to decide a particular matter” (para 59).
What that means, I suggest, is that the real disagreement between the judges is not over whether jurisdictional questions exist. It is on the extent to which courts should be willing to defer. The majority appears to be more comfortable with a broader use of deferential review, presumptively entitling home statute interpretation only to be reviewed deferentially. Neither Binnie J. nor Cromwell J.’s judgment would go that far. And this disagreement does not appear to be primarily a dispute about administrative law doctrine and jurisprudence; it is not a dispute about how to interpret Dunsmuir or other administrative law cases. It is, instead, about whether administrative decision-makers or courts are best situated to decide certain legal questions, about which allocation of responsibility will best protect the rule of law while respecting legislative intention and supremacy.
Is any one of the judgments obviously superior in this respect? I don’t think so. There are good reasons to presumptively defer to administrative decision-makers interpreting their home statutes – their expertise, their familiarity with the home statute and the factual context to which it applies, and their legislative mandate to engage in such interpretation. But there are also good reasons for a court to be cautious about presumptive deference on questions of law, even if arising in a home statute. One can be unduly romantic about the competence and expertise of administrative decision-makers to decide such questions; many administrative decision-makers have no legal training, and their expertise and may not be especially extensive. As Binnie J. said, not all administrative decision-makers are equally competent on questions of law.
That is not to say that the disagreement between the judgments is inconsequential; the three positions reflect fundamentally different conceptions about how courts can best ensure legitimacy in administrative decision-making. At the same time, however, each of the judgments is both justifiable and defensible; each can make different but plausible claims about how it ensures respect for rule of law and legislative supremacy.
What follows from that observation? On the one hand, it seems to suggest that the Court is unlikely to reach an agreement on these points in the foreseeable future. The differences between the judges are normative, not technical. On the other hand, if all of the approaches can be said to satisfy the objectives of administrative law, one can question the utility of the continuing debates that the Court has on these questions. I do understand the need for judges to remain true to what they believe is the principled and appropriate approach to substantive judicial review. But at the same time I would suggest that constant restatements of the law, and disputes about the best approach – the Court’s continual tweaking with how judicial review should occur – itself does a disservice to the law, and may do a greater disservice than simply picking one approach and sticking with it.
How can an administrative decision-maker, a client, a lawyer or even a lower court, predict with any certainty how the Supreme Court will respond to an administrative law case? And without that predictability, how can they advise clients or decide cases consistently with the Court’s jurisprudence? At some point I think the Court has to take greater responsibility for stating the law of judicial review with finality. To do otherwise is to sacrifice clarity and consistency to the search for perfection, and it’s hard to see what that search truly accomplishes.
With respect to the application of deference, I am surprised by the Court’s continued reluctance to acknowledge Binnie J.’s point that deference necessarily varies with the type of administrative decision being reviewed. The Dunsmuir test for reasonableness review directs courts to determine whether an administrative decision falls within the range of outcomes that were reasonably available to the decision-maker; it only makes sense to think that for some sorts of decisions the range of available outcomes will be much broader than in others. That is not to say that deference involves any standard other than reasonableness, but is only to describe what courts appear to do when assessing reasonableness in particular cases. Even in the Teachers’ Alliance case itself, when Rothstein J. purported to review the Commissioner’s decision with deference, he took considerable time to explain why the decision was justifiable – even perhaps “correct.” The reason for that seems to be, as was the case with the Court’s review of the labour arbitrator’s decision in Dunsmuir, that the decision was about the law, and the range of outcomes that could be justified was narrower than if, say, the decision was one of discretion or policy. Or, to put it slightly differently, on questions of law the Court is almost as well positioned to decide the matter as was the administrative decision-maker, and it will therefore only uphold the decision after turning its own mind to the best answer to give to that question, something which inevitably makes its review less deferential. By contrast, when a decision is one of, say, credibility of witnesses, the Court is incapable of meaningfully revisiting that decision, and is likely to review simply by determining whether the administrative decision-maker made any obvious errors in its analysis.
Be that as it may, on this point also the Court needs to identify a consistent approach. If rejecting Binnie J.’s perspective, then the Court must explain what it means when it says that deference is contextual, as Rothstein J. acknowledges it is. Otherwise in this aspect too administrative decision-makers, lower courts, lawyers and people affected by administrative decision-making will not have the consistent guidance to which they are entitled.