Double Aspect

I have previously mentioned Emmett Macfarlane’s Substack post written in response to my and Michael Plaxton‘s critiques of the Supreme Court of Canada for the inconsistency of its judgments on constitutional interpretation, as well as, in my case, other issues. Professor Macfarlane argues that we are too hard on the Supreme Court, and that such inconsistency as there is in its judgments is both unavoidable and not especially objectionable. I think all of this is wrong. In this post, I explain why. Professor Macfarlane also engages with an issue raised by Professor Plaxton, namely the role of the notwithstanding clause in all this. I mostly agree with what he says about that — indeed in some ways he is too soft on the notwithstanding clause enthusiasts — but I will defer this to a separate post.

Professor Macfarlane writes that the Supreme Court has never been especially consistent, or at any rate not for a very long time. He points to Reference re Secession of Quebec, [1998] 2 SCR 217, with its (rhetorical) emphasis on the “primacy” of constitutional text, and yet deployment of unwritten constitutional principles, as his main example. Professor Macfarlane makes this point later, but perhaps this would have been the better place to note that “some of these inconsistencies are likely less about individual judges contradicting themselves as they are about the consensual style of drafting on the Court”, which restrains individual judges from saying exactly what they mean or writing concurrences at every slight disagreement.

Professor Macfarlane insists that “we shouldn’t pretend the Court … could ever be, pristinely consistent in its approach to interpretation”. That’s just an academic pipe-dream. And indeed

it’s simply not unreasonable that a focus on the text might predominate in some instances and be subject to broader contextual or values-based readings in another, if only because in some provisions the text is broad and ambiguous and it others it is clear and specific.

Section 18 of the Constitution Act, 1867, at issue in  Alford v Canada (Attorney General), 2026 SCC 14, is of the latter sort, and the more textualist approach the Court took to it is not really inconsistent with “the dominant paradigm of living tree constitutionalism”. As Professor Macfarlane sees it,

the Court has tended to take a progressivist living tree approach to some Charter rights, is frustratingly tied to ancient and dubious precedent in federalism cases (hi Comeau!), while engaging in a frozen rights, caricature-of-originalism approach in Aboriginal rights cases.

In the end,

The law and precedent continue to constrain, and so long as the judges on our courts avoid becoming ideologues, and remain people of good faith and good effort to speak to the body of law we have, then we’ll be okay. 

To begin where Professor Macfarlane ends, a part, at least, of the reason why I have been writing my recent posts, and the reason for their vehemence, is precisely that I don’t see law and precedent constraining the Supreme Court. As I have shown, the majorities in Alford and in Taylor v Newfoundland and Labrador, 2026 SCC 5, simply ignore “law and precedent”, and it is difficult to see how authorities that are not even mentioned are constraining anyone. It is certainly open to the Supreme Court to distinguish a precedent, or even to overrule one. Provided that this is done openly and for plausible reasons (whether one agrees with them or not!), the precedent is still acting as a constraint on the court’s reasoning, if not also on its decision. But that just isn’t happening, and Professor Macfarlane’s stipulation to the contrary is, I am afraid, just that.

Of course it is true that no judge, let alone a multi-member court can be perfectly consistent in every case. As Lord Byron put it, “But if a writer should be quite consistent, How could he possibly show things existent?” But there are different sorts and levels of inconsistency, some of them tolerable, others not. The Secession Reference is an odd choice of exhibit in defence of the Supreme Court — it is, at best, a peculiar opinion formulated under difficult circumstances. This is how I tend to think of it, so to my mind, it is not crazy — though inartful — to say that while most constitutional disputes are to be resolved with reference to the text (“primacy”), some, like the one about secession, for which the text does not provide, are not. But a lot of serious people think that this opinion is a deplorable dog’s breakfast. Be that as it may, if everything the Supreme Court does is akin to the Secession Reference, even on my reading, we’re in a bad way. Fortunately, I wouldn’t go quite so far in my critique.

But, to repeat, inconsistency with precedent decided mere months ago — and not just inconsistency, but seeming obliviousness, as if the precedent didn’t exist — is a different issue from the perhaps inevitable (and still lamentable) logical defects of a judgment written by committee. There is an element of wilfulness about the former scenario that is (presumably) absent from the latter, and makes it far more damnable. This is not to defend the latter case, by the way; there are far too many Supreme Court decisions that are, indeed, incoherent because the Court attempted to manufacture a majority where none existed. Vavilov v Canada (Citizenship and Immigration), 2019 SCC 65, is of this sort, and I have condemned it for this (among other things). Still, there’s a sliver of an excuse there that simply isn’t available to cases like Alford.

And perhaps Professor Macfarlane recognizes as much, implicitly, because he attempts to discern a logic in the Supreme Court’s interpretation case law which would make it make some sense. Unfortunately, his suggestion that there is a distinction, perhaps even a sensible one, between Charter and federalism cases does not hold. For instance, the “living tree” metaphor, which he associates with the Charter cases, originates in Edwards v Attorney-General for Canada, [1930] AC 124 (PC) — a reference about the Constitution Act, 1867 —, and I think it is fair to say that one of its most celebrated invocations was in the discussion of the division of powers in Reference re Same-Sex Marriage, 2004 SCC 79. It was also invoked in  R v Comeau, 2018 SCC 15, which Professor Macfarlane associates with hidebound federalism. Conversely, Benjamin Oliphant and I have shown that various forms of originalist reasoning are on display in a number of Charter cases, and more recent cases, leading to Quebec (Attorney General) v 9147-0732 Québec inc, 2020 SCC 32, have been textualist if not originalist, as I explain in this piece.

In any case, the Supreme Court itself has never categorically said that the Charter had to be interpreted differently from the rest of the constitution. Although it occasionally speaks of “Charter interpretation”, it consistently cites non-Charter cases, including of course Edwards, when interpreting the Charter — just as it consistently cites Charter cases when interpreting other parts of the constitution. Taylor, in addition to Edwards, cites Caron v Alberta, 2015 SCC 56, the Secession Reference, and R v Blais, 2003 SCC 44. Conversely, Alford, though it ignores Taylor, cites Hunter v Southam Inc, [1984] 2 SCR 145 and Reference re Public Service Employee Relations Act (Alta.), [1987] 1 SCR 313.

Nor is the Supreme Court’s inconsistency a matter of treating clear provisions differently from vague ones. Again, the Court has never been explicit about there being such a distinction. But it is not obvious, to say the least, that the complex and careful wording of section 6 of the Charter, at issue in Taylor calls for more free-wheeling interpretation than the brief and general words of section 12, at issue in Quebec Inc. Yet it is in Taylor that the majority approached the text as if it hardly mattered, finding a right in the vibes of section 6 as a whole rather than its distinctive sub-sections, while in Quebec Inc it had been attentive to textual detail.

In short, Professor Macfarlane and some other defenders of the Supreme Court look at it through rose-tinted glasses. They suggest that the Court acts more rationally than it actually does, and that it is more constrained than it seems to be. The Supreme Court does not make a “good effort”, or seemingly any effort, to stay in touch with the law, in which case occasional failures should certainly be met with criticism, but also with a measure of understanding due to any human institution’s shortcomings. It declares the law one day only to ignore it the next, and disdains giving any explanations that anyone outside the institution could trust.

Contrary to what Professor Macfarlane suggests, this isn’t just a matter of academics pushing real-life institutions towards theoretical purity. For one thing, it was not academics who came up with the idea that “the law works itself pure”. It was an aspiration, to be sure, but an aspiration held by common lawyers over the centuries. The Supreme Court of Canada has abandoned it. And for another, the need for clarity, consistency, and ultimately predictability is not a mere theoretical concern, but a matter of great practical importance. It is Professor Macfarlane who has the luxury to be unbothered by judicial inconsistency because he “see[s] judging as inescapably political”. Canadian citizens and Canadian governments alike, and the lawyers who advise them, deserve better from their Supreme Court.