Jeffrey Pojanowski, whose contribution of “A View from South of the Border” to the Dunsmuir Decade symposium readers may recall, has posted a very interesting paper on “Neoclassical Administrative Law” on SSRN. (The article is to be published in the Harvard Law Review later this year.) Although written in an American context, Professor Pojanowski’s article should be read north of the border too, because it is framed around the tension that is central to Canadian, as well as if not more than, American administrative law: that between the Rule of Law and (what we in the Commonwealth call) Parliamentary sovereignty. Professor Pojanowski’s solution to this tension ought to be appealing in Canada ― though accepting it would require giving up some of the assumptions that are built into our administrative law.
Professor Pojanowski starts by describing three ways of addressing the conflict between the courts’ role of saying what the law is and the legislatures’ prerogative of committing certain governance issues to the resolution of administrative decision-makers. What he terms “administrative supremacy”
sees the role of courts and lawyers as limited to checking patently unreasonable exercises of power by the administrative actors who are the core of modern governance. To the extent that durable, legal norms are relevant, the primary responsibility for implementing them in administrative governance falls to the discretion of executive officials, who balance those norms’ worth against other policy goals. (7)
“Administrative skepticism”, by contrast,
rejects deference to agency interpretations of law, even if the agency is charged with administering the statute. Deference shirks the judicial duty to say what the law is and introduces a pro-government bias of dubious constitutional provenance. (14)
As for those cases where the lawyers’ traditional interpretive tools are of no avail, because the administrative decision-maker has been given a policy-making role, “the [American] administrative skeptic is more likely to recommend an approach that is both more radical and more modest: invalidating the provision on non-delegation grounds”. (16-17)
Finally, the approach Professor Pojanowski terms “pragmatist” “seeks to reconcile the reality of administrative power, expertise, and political authority with broader constitutional and rule-of law values”. (18) It is relatively deferential to administrative interpretations of law, but makes “certain exceptions, such as withholding deference on major questions or jurisdiction”, (18) and “may … demand evidence that the agency engaged in reasoned decisionmaking” (18) even on those issues where it is normally prepared to defer, both interpretive and policy ones.
In jurisprudential terms, administrative supremacy comports with “a form of legal realism that dissolves the line between legal interpretation and policymaking”, deeming “most interesting questions of legal interpretation … inextricable from legislative policy choices”. (13) The skeptical position embraces A.V. Dicey’s vision of ordinary courts interpreting law as the keystone of the Rule of Law. The pragmatist view reinterprets the Rule of Law as involving “requirements of fair participation and reasoned justification”, and asks the courts to enforce these requirements, rather than to impose their view of what the law actually is.
Professor Pojanowski articulates and begins the defence of another approach to administrative law, which differs from those just outlined, though it has some affinities with each of them, perhaps especially the skeptical one. This “neoclassical administrative law … is skeptical of judicial deference on questions of law but takes a much lighter touch on review of [administrative] agencies’ procedural and policymaking choices”. (23) It seeks to preserve, indeed it emphasizes, the distinction between law and policy, and makes the courts masters of the former while asking them to stay out of the latter.
In part, this is motivated by a “formalist” rejection of the “legal realist premise that all interpretive uncertainty involves policy choices calling for political accountability and non-legal expertise”. (27; footnote omitted) To be sure statutes sometimes employ language that is only amenable to policy-laden elaboration (such as “in the public interest”); such elaboration should be the preserve of administrative decision-makers, subject only to a thin rationality review. However, this is precisely because in such “cases … there is no surface upon which traditional lawyers’ tools can have purchase”, (31) and the obverse of accepting this is a denial of “the more generalized presumption of implicit [legislative] delegation of interpretive authority”, which is no more than “a legal fiction delicately veiling a functionalism that dare not show its face”. (26) Legal questions, even difficult ones that have “more than one reasonable answer”, (33) can and ought to be answered by the courts, although “reviewing judges are likely to confer at least some mild epistemic authority on expert agencies”. (25n) In addition, the “neoclassical” position rests on a belief in the importance of the legislation governing judicial review of administrative decisions, especially (in the United States) the Administrative Procedure Act.
But while the “neoclassical” approach is similar to the skeptical one in its confidence in the law’s autonomy from politics and policy, it does not go as far in its rejection of the administrative state. It does not seek to reinvigorate the constitutional non-delegation doctrine (which holds that only the legislature, and not its creatures in the executive branch, can make law). Instead, “[t]he neoclassical approach turns down the constitutional temperature”, (36) accepting that the administrative state’s rule-making and discretionary powers are here to stay. It, in other words, “classical Diceyan public law theory adapted and persisting in a new regulatory environment”. (38)
Professor Pojanowski ends by addressing some potential criticisms of “neoclassical administrative law”. Of greatest relevance to Canadians will be his admission that
much here turns on interpretive method. The extent to which appeal to craft determinacy is plausible goes a long way toward deciding whether neoclassicism is promising or misguided. Furthermore, if interpretive formalism is inferior to strong purposivism or dynamic statutory interpretation, the case for deference is far stronger. Those methods explicitly, and to a greater degree, call for interpreters to consider policy consequences and evolving public values alongside, and sometimes above, formalist tools. The more those values infuse legal interpretation, the stronger the bite of arguments for deference based on political accountability and technical expertise. (40; footnote omitted)
Professor Pojanowski points out, however, that the pragmatist view, at least, is also tenable only if there are legal answers to at least some interpretive questions, which its adherents exclude from the scope of judicial deference.
I find Professor Pojanowski’s summary of the various existing approaches to administrative law illuminating, and his own “neoclassical” approach, mostly compelling. As a matter of first principle, I might be attracted by anti-administrativist skepticism but, especially in Canada, it is not a plausible position. Whatever might be the persuasiveness of the originalist arguments in favour of the non-delegation doctrine, and of strict separation of powers more broadly, in the United States, I doubt one can take them far in Canada. Subject to (somewhat vague) constraints on legislative abdication, the delegation of discretionary and rule-making authority is within the powers of Parliament and the provincial legislatures under the Constitution Act, 1867. The question, then, is not whether we can burn the administrative state to the ground, but whether we can ensure that it remains subject to law. The “neoclassical” understanding of administrative law is a better way of doing that then the available alternatives.
At present, Canadian administrative law is torn between “administrative supremacy” and “pragmatism”. Dunsmuir v New Brunswick, 2008 SCC 9,  1 SCR 190, the soon-to-be-former leading case, is representative of the pragmatic approach, with its insistence that
[i]n judicial review, reasonableness is concerned mostly with the existence of justification, transparency and intelligibility within the decision-making process. But it is also concerned with whether the decision falls within a range of possible, acceptable outcomes which are defensible in respect of the facts and law. 
By contrast, cases such as Edmonton (City) v Edmonton East (Capilano) Shopping Centres Ltd, 2016 SCC 47,  2 SCR 293, which allow unjustified, unreasoned administrative decisions to stand in the name of an (almost?) irrebuttable “presumption of expertise”, epitomize administrative supremacy. That said, even the pragmatist strand of Canadian administrative law is infected with a metastasizing belief in the absence of legal answers to interpretive questions which in Dunsmuir and elsewhere has been said to warrant thoroughgoing deference to administrative interpretations of law.
In the circumstances, even reasserting the belief in the law is in fact autonomous from policy and politics, and that interpretive questions must be resolved by relying on legal rather than on administrative expertise, is a tall order. Professor Pojanowski points out that this belief goes hand in hand with a commitment to interpretation based “on the text’s original meaning, statutory context and structure, linguistic canons, and perhaps historical intent … rather than normative canons or legislative purpose at a high level of generality”. (34) Contrast this with the broad pro-regulatory purposivism of cases like West Fraser Mills Ltd v British Columbia (Workers’ Compensation Appeal Tribunal), 2018 SCC 22,  1 SCR 635, and you will see just how far we have to go. Yet West Fraser, with its purported acknowledgement of an “unrestricted delegation of power”  to an administrative tribunal, illustrates the dangers of the prevailing Canadian approach.
That said, I have a couple of interrelated concerns about Professor Pojanowski’s approach. The broader one has to do with judicial review of policy decisions, including “interpretation” (or rather construction) of such terms as “reasonable” or “in the public interest”. I am inclined to think that the approach to (constitutional) construction set out by Randy Barnett and Evan Bernick in “The Letter and the Spirit: A Unified Theory of Originalism” is apposite here. A reviewing court should ensure, not that just that the administrative decision is rational, but also that it is a good faith attempt to further the original purpose of the statutory provision on which it is based and of the statute as a whole. While legal craft may not be able to tell us how best to serve the public interest in a particular regulatory context, it can help shed some light on statutory purpose. Indeed, I think it is necessary that courts, rather than administrative decision-makers naturally incentivized to overvalue to importance of their perceived mission and to underrate the countervailing considerations that may well have led a legislature to limit their ability to advance their agenda, be the final arbiters of statutory purpose. As Justice Rand famously said in Roncarelli v Duplessis,  SCR 121
In public regulation … there is no such thing as absolute and untrammelled “discretion” … there is always a perspective within which a statute is intended to operate; and any clear departure from its lines or objects is just as objectionable as fraud or corruption. (140)
A related but more parochial concern has to do with constitutional law. Whatever deference might be warranted to administrative decision-makers engaged in the policy-laden elaboration of vague statutory terms, none should be accorded on constitutional issues. As a matter of the positive law of the Canadian constitution, the courts are the supreme arbiters of its meaning, against the executive branch as well as against the legislative. This question, if I understand correctly, does not even arise in the United States, but so long as Doré v Barreau du Québec, 2012 SCC 12,  1 SCR 395 remains on the books, it must be flagged in the Canadian context.
Another somewhat parochial question that the “neoclassical” approach to administrative law would force us to confront is that of what to do about a large number of statutory provisions that Canadian courts have so far more or less deliberately ignored or distorted beyond recognition. These are, on the one hand, “privative clauses” that purport to preclude review of administrative decisions; and on the other provisions such as section 18.1(4) of the Federal Courts Act, sections 58 and 59 of the Administrative Tribunals Act of British Columbia, and other provisions that seek to guide judicial review of administrative decisions. Privative clauses would be unconstitutional if taken literally; but instead of holding them unconstitutional and simply ignoring them as nullities, Canadian courts (used to) affect to take them seriously rather than literally, as indications that the decisions of tribunals protected by such clauses should be given greater deference. As the “presumption of deference” spread, even this position has become increasingly meaningless. Meanwhile, as co-blogger Mark Mancini has pointed out, in Canada (Citizenship and Immigration) v Khosa, 2009 SCC 12,  1 SCR 339, the Supreme Court subverted the guidance that section 18.1(4) provides, insisting on imposing its own views on the standard of review applicable to decisions of federal boards and tribunals. The Supreme Court has similarly ignored provisions creating statutory rights of appeal, treating appeals from administrative decisions like judicial reviews.
Professor Pojanowski calls for such legislation to be taken as binding law rather than guidelines to be subsumed into or overridden by the Suprme Court’s own views about judicial review. This should be the obvious thing to do: statute trumps the common law. However, there is a catch; two even. First, the principle of legality holds that common law rights, including the right to access courts, including, I think it is fair to say, for the purposes of judicial review, cannot be abolished by implication. I’m not sure whether this has repercussions for interpretation of legislation that guides judicial review, but it might in some cases. Often, however, the legislation is quite clear. Notably, section 58 of the above-mentioned BC statute requires review for patent unreasonableness, including on questions of law in the case of certain tribunals. I think the courts would need to squarely face, in an appropriate case, the question of whether legislatures are constitutionally permitted to set the bar so high. And the courts should stop pretending to attach any significance to unconstitutional privative clauses.
Professor Pojanowski has articulated an approach to administrative law that is at once principled and (relatively) realistic. It responds to concerns that animate not only American, but also Canadian law, and should therefore be of considerable interest to us, not just as a comparativist curiosity, but as a source of compelling ideas. For this approach to take hold in Canada, long-held assumptions will require revision, and difficult questions will need answering. Yet it is quite clearly superior to available alternatives. Count me a cautious neoclassicist.