Advocates for the Rule of Law
Earlier this month, Mark Mancini wrote two very thoughtful blog posts on the Double Aspect blog, attempting to bring administrative law back to first principles. These intriguing posts are worthy of commentary. I will respond to Mancini’s two posts today, and follow up next week with an addendum in light of the Supreme Court of Canada’s recent decision in Canada v Chhina.
Historical Justifications for Deference
Much of modern Canadian administrative law concerns the question of “standard of review”: how much, if any, deference, an administrator should be given in interpreting the law. Mancini goes into a deep dive into the history of how we got here: a lamentation of courts being an (alleged) impediment to the realization of government policy goals. Mancini labels these critiques, leveled primarily from the 1930s through 1970s, as “Progressive”, referring to the New Deal-era term.
I have perhaps more sympathy than Mancini with respect to such scholars – to say nothing of democratically elected legislators – being exasperated by questionable judicial interpretations of statutes. The “Lochner era” in American constitutional history is rightly criticized, in my opinion, for judicial overreach. Even in Canada, courts were infamous for interpreting labour legislation in particular, seeming to stretch the definition of a decision made “without jurisdiction”.
But Mancini astutely notes that, regardless of historical context, the administrative state is vastly different in the modern era. While labour regulation was the Progressives’ quintessential point of reference, the modern administrative state addresses matters including revocation of citizenship, entitlement to refugee status, eligibility for social support, determinations of detention standards, and many others. Administrative decision-makers frequently have no legal training. It seems problematic to base the modern law of judicial review in concerns about 1970s judicial overreach in the labour law context.
Mancini cites three reasons cited by modern and historic defenders of deference. The first is expertise. Expertise may sometimes be a policy reason to defer to a decision-maker. But as Mancini and others have noted, it is frequently not present (Vavilov is a great example). Expertise seems uncompelling as a stand-alone legal reason for deference. It remains the legislature’s decision to prescribe deference in a particular statute as a matter of policy – not the purview of the courts to declare it to be so in all contexts.
The second alleged justification for deference is efficiency. Mancini shows that this is frequently not present in administrative decision-makers. I take his examples of poorly functioning administrators. A comprehensive analysis would arguably be necessary to determine whether, in the aggregate, administrators are more efficient than courts. To be sure, courts are hardly paradigms of efficiency. Nonetheless, even if administrators make decisions more efficiently than courts, this cuts both ways. Why? Because a court reaching a determinative interpretation of a statute allows administrators to make subsequent decisions more efficiently, not needing to re-interpret the law from a blank slate. It also allows parties to order their affairs more predictably, lessening the need to resort to the administrator in the first place.
The third alleged reason for deference is independence. Mancini doubts the presence of independence as an empirical fact in many administrators. I suspect this is very administrator-specific, but Mancini points out enough instances of an obvious lack of independence so as to cast doubt on the wisdom of using independence as a stand-alone reason for deference. (In any event, there are circumstances where a lack of independence can raise procedural fairness concerns, causing one to wonder what role independence should play in the substantive review analysis.)
Ultimately, Mancini builds a strong case undermining the policy reasons for deference to administrators, both historically and contemporarily. I do regret the emphasis on the term “Progressive” as the purported justification for deference, as this implies the political left should be inclined to support deference to administrators while the political right should not. From an institutional design perspective, we wish to remove such political considerations. And to be sure, Mancini is clear he is referring to the “Progressive” movement of a particular era. In any event, as I argue below, deference can be an impediment to the realization of small-p progressive results.
When to Defer: Beyond Standard of Review
The question of when to defer and when not to, as a matter of political theory, is debatable. It seems as though factors in favour of deference may be present in the case of many administrators’ decisions, but not others. But from a legal perspective, Mancini convincingly notes that determining how much deference is due in each case is a time-consuming and usually unwarranted exercise. A “rule” rather than a “standard” appears appropriate given the ubiquity of judicial review.
In the face of this, Mancini proposes that questions of law be reviewed on a correctness standard, unless particular legislation indicates that deference is warranted. A statute would typically suggest that correctness should not be applied through a privative clause.
I find this proposal attractive for numerous reasons. First, the statute gives the administrator its power. The courts – tasked with statutory interpretation in our system of governance – should determine that an administrator stays within its bounds in this area. As Justice Rowe wittily noted last year in West Fraser, working day-to-day with a statute’s subject matter does not bestow expertise in statutory interpretation, which is a question of legal analysis. Second, correctness review has the advantage ensuring of consistent interpretations of law, allowing parties to order their affairs. Third, and most importantly, such an approach respects legislative intent, which courts are constitutionally obliged to respect and is inherently statute-specific. As I noted when discussing the 1979 CUPE case on Double Aspect in December, the language of “standard of review” is found in neither common law legal tradition nor most statutory texts. It has caused mischief in Canada. Simply asking whether a statute calls for deference is a straight-up task in statutory interpretation for which courts are eminently qualified.
If a legislature wishes to insulate an administrator’s decision from judicial review, it is welcome to do so within constitutional constraints. This will necessitate a body of law determining whether deference is called for by any particular statute. But it will probably not be complicated in the vast majority of cases, with the presence or absence of a privative clause likely being determinative of the matter. If deference is appropriate, an administrator’s decision should be upheld unless indefensible. In very rare cases, the rule of law may mandate that courts intervene despite a privative clause if the administrator has come to an indefensible interpretation of the statute. But this is the only real circumstance that the rule of law and parliamentary supremacy come into tension – and it is rare. (This puts aside the question of whether certain matters such as constitutional interpretation ipso facto call for no deference. Another possible instance of this, identified by my soon-to-be colleagues at Robson Hall, Amar Khoday and Gerald Heckman, is the interpretation of international human rights law.)
It may seem that a “correctness unless the statute says otherwise” standard of review excessively complicates what could be a simple analysis through a single standard of reasonableness. Courts will likely need to interpret each statute to determine whether deference is due. But courts find themselves practically interpreting each statute in any event to determine whether particular interpretations are reasonable. The fact is that each statute is unique and operates in different circumstances. Having a system of administrative law that recognizes as much appears eminently reasonable.
Of course, none of this prevents common sense “deference” in circumstances where an administrator truly is expert. But this is not deference that is due de jure. For statutes such as the Criminal Code, the Class Proceedings Act, or the Income Tax Act, appellate courts do not defer to “expert” Provincial Court, Superior Court, or Tax Court judges who may know more about the area. Although consideration of those judges’ “expertise” may rightly caution an appellate judge against coming to a contrary interpretation of the statute, this is not deference that is due as a matter of law but rather as a matter of common sense. To quote Justice Scalia:
If I had been sitting on the Supreme Court when Learned Hand was still alive, it would similarly have been, as a practical matter, desirable for me to accept his views in all of his cases under review, on the basis that he is a lot wiser than I, and more likely to get it right. But that would hardly have been theoretically valid. Even if Hand would have been de facto superior, I would have been ex officio so. So also with judicial acceptance of the agencies’ views.
I believe Justice Scalia overstated the matter. Legislatures have the right to promulgate the law and I fail to see why, in principle, they cannot use their judgment to delegate law-making or law-applying authority to expert administrators. But the legislature actually has to use said judgment to give said administrators such power. It ought not to be assumed.
Concluding Thoughts
Mark Mancini’s recent thoughtful contributions to the discussion surrounding standard of review should be welcomed. His suggestion for a new rule to govern this area, even if not one typically proposed, is principled. While some political connotations surrounding the debate are unfortunate, we can see that these issues of administrative law and standard of review affect such vast areas of legal practice and the role of government that attempts to boldly depart from legal tradition should be treated with caution.
