Double Aspect

My last several posts have been concerned with the Supreme Court of Canada’s failure not only to follow but even to acknowledge its own recent precedents on constitutional interpretation, first in Taylor v Newfoundland and Labrador, 2026 SCC 5 and then in Alford v Canada (Attorney General), 2026 SCC 14. I have argued that this repeated failure does not reflect well on the Supreme Court, to put it mildly. However, since writing my post on Alford, a counterargument has occurred to me. In this post, I will examine and — spoiler alert! — reject it.

This objection is based on a paper Paul Daly: “The Signal and the Noise in Administrative Law“. Professor Daly argued that, for reasons I shall presently explain, only some of the Supreme Court’s administrative law decisions should be taken seriously as decisions about administrative law (i.e. as the signal in his title). Others, though superficially relevant, are really best ignored from an administrative law standpoint (the noise). Professor Daly’s argument is focused on administrative law, but might it be more broadly applicable? Could something similar be happening in the law of constitutional interpretation? I think it’s a thought worth considering, but I am not persuaded.

The reason for the existence of signal and noise cases in administrative law, Professor Daly suggests, is that administrative law is a complicated thing. You might even say it has a double aspect. (That’s my bad joke. Don’t blame Professor Daly.) He writes:

Administrative law consists of general principles that have to be applied to different areas of substantive law. … These general principles exist at one remove from the substantive law that provides the context in which administrative decisions are taken. As a result, the Court may grant leave to appeal (and may eventually decide a case) for one of two reasons: it may wish to answer a question or questions relating to the general principles of judicial review; it may wish to answer a question of substantive law; some combination of the two is also possible.

Professor Daly thinks that more often than not, it is substantive law, not the administrative law principles, that the Court is really interested in.

When the Supreme Court decides a case which it has chosen for its potential to provide guidance on substantive law, it will still say something about the general principles of administrative law. But it will do so with less care, and with less intentionality, than if and when these principles are its actual focus and reason for taking up the case. The result is noise: cases where the Court seems to address administrative law issues, but which don’t really matter to it from that perspective, and which end up only distracting the scholars and practitioners who are interested in general principles rather than in substantive law.

Now, I think this duality may be especially strong in administrative law, but it is not unique to it. In constitutional cases too, there are general principles at play, and substantive law. The principles of proportionality analysis in rights cases are examples of the former. The actual rights — freedom of expression, freedom of religion, etc — of the latter. In deciding a constitutional case, the Supreme Court might be interested in one or the other, or both, though I wonder whether that emphasis exists from the moment the Court grants leave to appeal or develops in the course of argument or indeed of the drafting of the opinion(s).

Arguably, the rules of constitutional interpretation are also among the general principles of constitutional law, and thus only an occasional focus of the Supreme Court’s interest, so that some significant part of what it says about them is noise rather than signal. I’m not quite sure about this. Interpretive rules are, in a sense, constitutive of the substantive law: what the scope and meaning of a Charter right is turns on the way in which the text entrenching it is interpreted, whereas that’s not the case of, say, the rules of immigration law and the administrative law standard of review, or even the way in which that Charter right’s limitations are assessed (though someone like Grégoire Webber might disagree on the latter point). But still, it may well be possible to treat a case as being primarily about the substance of a constitutional rule, for example because all plausible interpretive methods converge, if not on a single interpretation of a provision, then on interpretations that overlap on the point at issue. In other cases, this will not be true, and interpretive methods will have to come into focus.

If the signal/noise framework applies, then how do we distinguish the two? Professor Daly has several suggestions. First, the more space a judgment devotes to discussing general principles as opposed to substantive rules, the more it can be taken to provide a clear signal; short discussions are likely to be noise. While Professor Daly discusses it separately, I think his suggestion that signal cases are more likely to consider a lower court’s treatment of the issue than the noise ones is of a piece with this. Next, “the presence of concurring and dissenting reasons as to a majority judgment’s treatment of the general principles … will suggest that a decision should be treated as signal rather than noise”. (78) Professor Daly doesn’t say anything about what we should make of unanimity on a point of general principle, but my own hunch is that it should not count either way. Lastly, “a decision which is out of line with other decisions on the general principles of administrative law is more likely to be noise than signal”. (80)

How does this apply to cases on constitutional interpretation? It is tempting to dismiss Alford as noise: the discussion of the “principles of constitutional interpretation” is short (and unanimous, for whatever that is worth), and, as I argued in my last post, out of kilter with recent precedent. But the difficulty is that this recent precedent, and the recent precedent before that, do not fit the framework so neatly.

Taylor looks like signal in some key ways: the discussion of constitutional interpretation there is detailed and extensive, and it is precisely interpretation that divides the judges, especially the majority and the two minority judgments. But, as I have argued here and here, it too is out of kilter with precedent, notably R v Comeau, 2018 SCC 15 and, especially, Quebec (Attorney General) v 9147-0732 Québec inc, 2020 SCC 32 and the line of cases that led to it.

And Quebec Inc itself very much looks like a signal case: again, it features a lengthy discussion of interpretive principles and a sharp division between the majority and the concurrence on precisely this point. Indeed there is good reason to think that the concurrence was meant to be a majority or even unanimous opinion, and that its treatment of constitutional interpretation is what lost enough support for a very differently written opinion to emerge as the majority one. But Quebec Inc, or at least the (relatively short) line of cases leading to it, are also very different from any number of past decisions (I have more on this in this detailed study).

In short, although it has some intellectual attraction in this context, Professor Daly’s signal/noise framework simply cannot cope with the wild swings of the Supreme Court’s case law on constitutional interpretation. (Needless to say, this is no criticism of the framework as applied to administrative law, or of Professor Daly, who only meant it to be applied there.) If multiple cases decided at fairly short intervals are all meant to provide clear signals about constitutional interpretation, then wild inconsistencies between them defeat that purpose entirely, and we are left with the conclusion that all there is to the law of constitutional interpretation in Canada is noise.

This is not to suggest that a situation where signal and noise coexist is an especially desirable one. Administrative law (excluding the treatment of administrative decisions that affect Charter rights), is a model of stability in comparison to constitutional law, what with its once-a-decade demolition-and-reconstruction cycles. Nonetheless, Professor Daly argued that it too requires a more coherent approach that would avoid the need for lawyers and lower court judges to parse Supreme Court decisions to identify the ones, or the parts of them, that meant something. This is obviously right (however much I may disagree with the substance of Professor Daly’s proposals, which were to a considerable extent adopted in  Vavilov v Canada (Citizenship and Immigration), 2019 SCC 65). 

In a Substack post, Emmett Macfarlane suggests that this is impossible to achieve in constitutional law, which always has been and is bound to remain eclectic. People like Michael Plaxton and I are, therefore, wrong to come down so hard on the Supreme Court for its inconsistencies. The post deserves a more detailed response, which I will try to write sometime soon, but in the meantime let me just note that, as I have argued in this piece, the requirements of the rule of law, including clarity and predictability, ought to apply in constitutional law as much as else. It is fair to say, as Lon Fuller does, that these requirements are not absolute and cannot be perfectly realized. But, as I have shown in my posts on Taylor and Alford, the Supreme Court is not even trying. That’s not good enough, and there is no excuse for it.