Does the Constitution Mean Anything?

In defence of textualism in constitutional interpretation

The Stereo Decisis podcast recently devoted an episode to a discussion of a case that I have covered here, 9147-0732 Québec inc v Directeur des poursuites criminelles et pénales, 2019 QCCA 373, in which the Québec Court of Appeal held that corporations could avail themselves of the protection of section 12 of the Canadian Charter of Rights and Freedoms against “cruel and unusual treatment and punishment”. While the hosts Robert Danay, Oliver Pulleyblank, and Hillary Young disagreed on the merits of the issue before the court, they were, I take it, agreed on one thing: the approach to interpreting section 12 on which my post relied is not compelling. And indeed my post was pointedly textualist, and intended as a bit of a provocation to the adherents to Canadian consensus approach to the constitution, which is anything but. I am glad that it worked, and that we are, as a result, having a bit of a debate on constitutional interpretation; and all the more so since, in the course of this discussion, my critics nicely expose the weakness of their position.

Briefly, I had argued that section 12 does not apply to corporations because the word “cruel” refers to the wilful infliction of or indifference to pain or suffering, and pain or suffering is something that corporations are not capable of. I added a discussion of the evolution of the provisions intended to limit punishments from the Magna Carta, to the Bill of Rights, 1688 and the Eighth Amendment to the US Constitution, to section 12, during the course of which the prohibition on “excessive fines” (to use the language of the Bill of Rights) fell by the wayside and was left out of the Charter. Considerations about whether it would have been a good idea for the Charter’s framers to have made a different choice and included a protection against excessive fines, which in effect is what the Québec Court of Appeal decided, are in my opinion irrelevant.

The hosts of Stereo Decisis took issue with that. We just can’t interpret the Charter simply by looking at what it says. Mr. Pulleyblank insisted that “‘[c]ruel and unusual’ is a bad phrase. It doesn’t really mean cruel and it doesn’t really mean unusual.” And beyond this particular provision, Professor Young said that the Charter is written in “rather loose language”, so that answers to questions about its meaning “can’t be found in the words”. Rather, they can only be obtained by asking what the Charter ought to mean. “You have to look beyond the words”, to “normative” considerations, such “how you feel about the Charter versus legislative authority”. The Québec Court of Appeal, for instance, had to decide whether “this particular right should apply to corporations”. (Emphasis mine) And that decision can yield, as Mr. Pulleyblank put it, “a norm that is different than either of those words [cruel and unusual] or both of those words together”.

Normative considerations are what caused the hosts to disagree about the outcome of the case. Mr. Danay said that “[w]e ought not to try to limit Charter rights. … If the Charter seems like it could protect something, probably a better reading … would be to protect that thing.” Professor Young, by contrast, saw a greater role for deference “to legislation enacted by elected legislatures”, and added that “[i]f we were talking about human beings’ rights, I would be less inclined to interpret so narrowly but I’m not super sympathetic about arguments for corporations’ rights against cruel and unusual treatment”. It was, as Mr. Pulleyblank summed it up, “just a disagreement” about “the impact on the democratic process”.

In my view, the hosts’ criticism of my textualist interpretation are weak, and their own approach grounded in vague normative considerations, unattractive. Now, it’s important to understand what textualism is not, and what it is. No textualist, for example, would say that answers to all constitutional questions can be found in the words alone. Sometimes, it is indeed necessary to go beyond the words of a provision. Some words that the Charter‘s framers used are vague. Context can clarify what at first glance appears to vagueness; in other cases, it might tells us that the most straightforward reading of a word whose import at first seems clear is not the most accurate one. Thus, contrary to what Mr. Pulleyblank rather derisively implied, my “going beyond the text” to look at section 12’s historical antecedents does not make me a bad textualist. Textualism is, in short, the idea that constitutional text, read in its proper context, binds ― insofar as it has an ascertainable meaning; it is not the view that text alone will always answer all constitutional questions. (In any case though, my ultimate commitment is to public meaning originalism, which starts, but does not always end, with textualism.)

So textualism can acknowledge the vagueness of a constitutional provision, but it will insist on not merely stipulating that its language is “bad” or “loose”, or that, if it is somewhat vague, it is incapable of providing any real guidance to the interpreter. The word “cruel”, in section 12, is a nice illustration. Of course, it is vague to a considerable extent. No amount of looking at dictionaries will tell us whether, say, a parole ineligibility period longer than an offender’s life expectancy is cruel (the main question in R c Bissonnette, 2019 QCCS 354) and, as a public meaning originalist, I do not think that knowing how the Charter‘s framers would have answered that particular question tells us much about the meaning of section 12 either. But it doesn’t follow that the word cruel is poorly chosen or that it has no real import at all. In the case before the Québec Court of Appeal, looking at the word’s ordinary meaning was helpful, indeed sufficient to dispose of the dispute (which an examination of the context confirmed).

The Stereo Decisis hosts never actually disputed this ― they did not refer to definitions of the word “cruel” that contradicted the claims that Justice Chamberland (who dissented at the Court of Appeal) and I made about it. (At least that’s how I understood them; as I was writing this post, Benjamin Oliphant suggested that “the hosts raise a worthwhile challenge to [my] interpretation of section 12. What if the words ‘cruel and unusual’ are properly understood to mean “grossly disproportionate’ … ?” I don’t think the hosts said that section 12 actually means this ― only that it has been read in this way by the Supreme Court. And I don’t think that “cruel” actually means “grossly disproportionate”. Again, dictionary definitions tend to emphasize wilful infliction of pain. Moreover, section 12 applies not only to “punishment” but to other “treatment” of the individual by the state. While it makes sense to speak of cruel treatment, I don’t think that “grossly disproportional” works here; disproportional to what?)

As I understood the Stereo Decisis hosts, they took what I can only describe as a dogmatic position that a word like “cruel” must be so vague as to provide no guidance. I don’t think that going into an interpretive exercise with a pre-determined view of this sort is right. Vagueness is not an all-or-nothing thing; a word, or a provision, can be vague as to some questions but not others. The interpreter needs to make a reasonable effort to glean what guidance can be had from the text and context before concluding that they “run out” and that the question facing him or her must be answered by looking at other considerations.

And then, the interpreter needs to face the question of what considerations should be looked at when, and to the extent that, a constitutional provision does run out. (In originalist terms, this is the question of what theory of construction one must adopt for those cases that interpretation does not settle.) The Stereo Decisis hosts suggest that we must go straight to very general normative views about the Charter and legislative power. As their discussion shows, however, this approach is not especially fruitful, in that it promptly leads to stark normative disagreement between those who would maximize the scope of the Charter‘s limits on government power and those who would reduce it in the name of preserving legislative authority. The two sides of this dispute have little to say to one another; both argue that the case should simply be decided by following their normative priors; they can only count heads to see who wins on any particular panel. Adjudication along these lines is not readily distinguishable from a legislative power struggle.

I do not mean to deny that cases where a court can do no better may arise from time to time. Still, I think that we should be uneasy about this prospect. Telling judges that it’s normal, rather than exceptional and worrying, for them to decide constitutional cases by reference to their own normative commitments produces nefarious consequences, as judges come to think that their personal understanding of right and wrong is more important than the law. From constitutional cases, this belief bleeds into other areas of the law ― into cases of ordinary statutory interpretation and even common law ones. This destroys the Rule of Law and removes the most important constraint on judicial power, which is the requirement to (normally) follow the law, be it constitution, statute, and precedent, that someone else has first set out.

Moreover, if constitutional disputes can only be decided by reference to what are political rather than legal considerations, then it is not obvious, as a normative matter, why they should be decided by the courts rather than by political institutions. (This is, of course, especially true of cases that involve individual rights; federalism disputes arguably require a neutral arbiter, but even there, it is not quite clear why the arbiter should be judicial in character.) And, as a descriptive matter, those who hold to the view that constitutional texts are more or less meaningless don’t even have access to the positive law argument I have made here that, as a textual matter, our constitution actually requires judicial supremacy. They must attempt to answer the question of whether it does so with normative arguments alone, and are unlikely to convince anyone not predisposed to agree with them.

It is much better, as well as more consistent with our Rule of Law tradition and with the positive law of our constitution, to insist that judges ascertain the meaning of the law given them, and if the meaning does not resolve the dispute they have to settle, that they endeavour to implement this law, not on the basis of their predilections, but of the law’s purposes. A judge who happens to share my distaste for most economic regulation can and should nevertheless conclude that, while an additional obstacle to such regulation’s excesses in the shape of an extension of the scope of section 12 to corporations would be normatively desirable, the constitution that we actually have does not raise this obstacle in the grasping legislatures’ way. But for him or her to be able so to conclude, that judge must be committed to elucidating and applying the law, instead of believing that judicial office gives one carte blanche to implement one’s own preferences.

Constitutional interpretation isn’t discussed enough in Canada. A general lack of interest, caused by overconfidence in a living constitutionalist orthodoxy, has meant that when Canadian lawyers confront questions of constitutional interpretation they are liable to reason in ways that are not compelling. Sadly, the Stereo Decisis discussion of the question whether section 12 of the Charter extends to corporation was illustrative. It relied on a mistaken assumption that constitutional language is infinitely malleable, with the result that, as Mr. Pulleyblank put it, “[i]f you want to go one way you can probably get there. If you want to go the other way you can probably get there.” Descriptively, this mischaracterizes our constitutional documents, which are rather less vague than is sometimes said. Normatively, a state of affairs in which constitutional law dissolves in competing assertions about the appropriate relationship between courts and legislatures, has little to recommend itself.

These two defects feed into each other. The less faith one has in the capacity of constitutional texts to guide their interpreters, the more power one is inclined to grant judges (even if only to seek to claw it back through free-floating doctrines of deference). The less one accepts limits on judicial power, the more one is tempted to see vagueness in every text, without seriously examining it. Still, I hope that, by discussing constitutional interpretation and calling into question beliefs about it whose truth has too long simply been assumed we will make much needed progress.

Anti-Bullying Law Struck Down

Last week, the Supreme Court of Nova Scotia struck down the province’s recently-enacted anti-cyber-bullying legislation, the Cyber-Safety Act. In Crouch v. Snell, 2015 NSSC 340, Justice McDougall holds that the Act both infringed the freedom of expression protected by s. 2(b) of the Canadian Charter of Rights and Freedoms, and made possible deprivations of liberty inconsistent with the principles of fundamental justice, contrary to s. 7 of the Charter. In this post, I summarize Justice McDougall’s reasons. (At great length, I am afraid, partly because it is important to explain the somewhat complicated legislation at issue, and mostly because the opinion covers a lot of constitutional ground.) I will comment separately.

Although laws against cyber-bullying are often justified by the need to protect young persons (especially children) from attacks and harassment by their peers, the parties in Crouch were adults, former partners in a technology start up who had had a falling out. Mr. Crouch alleged that “Mr. Snell began a ‘smear campaign’ against him on social media.” [22] Mr. Crouch eventually responded by applying for a “protection order” under the Cyber-Safety Act.

The Act, whose stated “purpose … is to provide safer communities by creating administrative and court processes that can be used to address and prevent cyberbullying,” (s. 2) makes it possible for persons who consider that they are being the victims of cyber-bullying (or for their parents and police officers, if they are minors) to apply for an order that can include prohibitions against its target communicating with or about the applicant, or using specified electronic services or devices. The Act defines cyberbullying as

any electronic communication through the use of technology including, without limiting the generality of the foregoing, computers, other electronic devices, social net works, text messaging, instant messaging, websites and electronic mail, typically repeated or with continuing effect, that is intended or ought reasonably [to] be expected to cause fear, intimidation, humiliation, distress or other damage or harm to another person’s health, emotional well-being, self-esteem or reputation, and includ[ing] assisting or encouraging such communication in any way. (Par. 3(1)(b))

While some earlier cases read this definition as including requirement of malice into this definition, Justice McDougall considers that it included not only actions that had a “culpable intent” but also “conduct where harm was not intended, but ought reasonably to have been expected.”[80]

The applications are made “without notice to the respondent.” (Subs. 5(1)) If “the justice determines, on a balance of probabilities, that … the respondent engaged in cyberbullying of the subject; and … there are reasonable grounds to believe that the respondent will engage in cyberbullying of the subject in the future,” (s. 8) he or she can issue a “protection order.” Once an order is granted by the justice of the peace, it must be served on its target. A copy is forwarded to the Supreme Court, where a judge must review the order and confirm it (with or without amendment) if he or she “is satisfied that there was sufficient evidence … to support the making of the order.” (Subs. 12(2)) If the judge is not so satisfied, he or she must “direct a hearing of the matter in whole or in part,” (Subs. 12(3)) at which point the target of the order as well as the applicant are notified and can be heard.

Mr. Crouch’s application resulted in a protection order being granted by a justice of the peace. Reviewing it, Justice McDougall finds that some of Mr. Crouch’s allegations were unsupported by any evidence; indeed, in applying for the protection order, Mr. Crouch misrepresented a perfectly innocent statement made by Mr. Snell as a threat by taking it out of the context in which it had been made. Nevertheless, there was enough evidence supporting Mr. Crouch’s complaint for Justice McDougall to confirm, in somewhat revised form, the protection order that prohibited Mr. Snell “from directly or indirectly communicating with” or “about” Mr. Crouch, [23] and ordering him to remove any social media postings that referred to Mr. Crouch explicitly or “that might reasonably lead one to conclude that they refer to” him. [73] This confirmation was subject to a ruling on the Cyber-Safety Act‘s constitutionality, which Mr. Snell challenged.

His first argument was that the Act infringed his freedom of expression. Remarkably, the government was not content to argue that the infringement was justified under s. 1 of the Charter, and actually claimed that there was no infringement at all, “because communications that come within the definition of ‘cyberbullying’ are, due to their malicious and hurtful nature, low-value communications that do not accord with the values sought to be protected under s. 2(b).” [101] Justice McDougall rejects this argument, since the Supreme Court has consistently held that “[t]he only type of expression that receives no Charter protection is violent expression.” [102] In finding that both the purpose and the effect of the Act infringed freedom of expression, Justice McDougall cites Justice Moir’s comments in Self v. Baha’i, 2015 NSSC 94, at par. 25 :

[a] neighbour who calls to warn that smoke is coming from your upstairs windows causes fear. A lawyer who sends a demand letter by fax or e-mail causes intimidation. I expect Bob Dylan caused humiliation to P.F. Sloan when he released “Positively 4th Street”, just as a local on-line newspaper causes humiliation when it reports that someone has been charged with a vile offence. Each is a cyberbully, according to the literal meaning of the definitions, no matter the good intentions of the neighbour, the just demand of the lawyer, or the truthfulness of Mr. Dylan or the newspaper.

(Self was the case where the judge read a requirement of malice into the definition of cyber-bullying. There had, however, been no constitutional challenge to the Cyber-Safety Act there. Incidentally, Self also arose from a business dispute.)

The more difficult issue, as usual in freedom of expression cases, is whether the infringement is a “reasonable limit[] prescribed by law that can be demonstrably justified in a free and democratic society,” as section 1 of the Charter requires. In the opinion of Justice McDougall, the Cyber-Safety Act fails not only the Oakes test for justifying restrictions on rights, but also the requirement that such restrictions be “prescribed by law.”

Mr. Snell argued that the definition of cyber-bullying in the Cyber-Safety Act was too vague to count as “prescribed by law.” Justice McDougall considers that the definition “is sufficiently clear to delineate a risk zone. It provides an intelligible standard” [129] for legal debate. However, in his view, the same cannot be said of the requirement in section 8 of the Act that there be “reasonable grounds to believe that the respondent will engage in cyberbullying of the subject in the future.” Justice McDougall finds that “[t]he Act provides no guidance on what kinds of evidence and considerations might be relevant here [and thus] no standard so as to avoid arbitrary decision-making.” [130] While risk of re-offending is assessed in criminal sentencing decisions, this is done on the basis of evidence, rather than on an ex-parte application that may include only limited evidence of past, and no indication of future, conduct. Here, “[t]he Legislature has given a plenary discretion to do whatever seems best in a wide set of circumstances,” which is likely to result in “arbitrary and discriminatory applications.” [137]

Although this should be enough to dispose of the case, Justice McDougall nevertheless goes on to put the Cyber-Safety Act to the Oakes test. He concludes

that the objectives of the Act—to create efficient and cost-effective administrative and court processes to address cyberbullying, in order to protect Nova Scotians from undue harm to their reputation and their mental well-being—is [sic] pressing and substantial. [147]

However, he finds that the ex-parte nature of the process created by the Cyber-Safety Act is not rationally connected to these objectives. While proceeding without notice to the respondent may be necessary when the applicant does not know who is cyber-bullying him or her, or in emergencies, the Act requires applications to be ex-parte in every case. It thus “does not specifically address a targeted mischief.” [158]

Nor is the Act, in Justice McDougall’s view, minimally impairing of the freedom of expression. Indeed, he deems “the Cyber-safety Act, and the definition of cyberbullying in particular, … a colossal failure” in that it “unnecessarily catches material that has little or nothing to do with the prevention of cyberbullying.” [165] It applies to “both private and public communications,” [165] provides no defences ― not even truth or absence of ill-will ―, and does not require “proof of harm.” [165]

Finally, Justice McDougall is of the opinion that the positive effects of the Cyber-Safety Act ― of which there is no evidence but whose existence he seems willing to “presume[]” [173] ― do not outweigh the deleterious ones. Once again, the scope of the definition of cyber-bullying is the issue: “[i]t is clear that many types of expression that go to the core of freedom of expression values might be caught” [175] by the statute.

In addition to the argument based on freedom of expression, Mr. Snell raised the issue of s. 7 of the Charter, and Justice McDougall addresses it too. The Cyber-Safety Act engages the liberty interest because the penalties for not complying with a “protection order” can include imprisonment. In Justice McDougall’s view, this potential interference with liberty is not in accordance with the principles of fundamental justice ― quite a few of them, actually. The ex-parte nature of the process the Act sets up is arbitrary, since as Justice McDougall already found, it lacks a rational connection with its objective. The statutory definition of cyber-bullying is overbroad, for the same reason it is not minimally impairing of the freedom of expression. The “requirement that the respondent be deemed likely to engage in cyberbullying in the future is incredibly vague.” [197] Moreover, “the protection order procedure set out in the Cyber-safety Act is not procedurally fair,” due mostly to “the failure to provide a respondent whose identity is known or easily ascertainable with notice of and the opportunity to participate in the initial protection order hearing.” [203] Finally, Justice McDougall adopts Justice Wilson’s suggestion in R. v. Morgentaler, [1988] 1 S.C.R. 30, that a deprivation of a s. 7 right that is also an infringement of another Charter right is not in accordance with the principles of fundamental justice. The Cyber-Safety Act infringes the freedom of expression, which “weighs heavily against a finding that the impugned law accords with the principles of fundamental justice.” [204] As with the infringement of the freedom expression, that of s. 7 is not justified under section 1 of the Charter.

As a result, Justice McDougall declares the Cyber-Safety Act unconstitutional. The statutory scheme is too dependent on the over-inclusive definition of cyber-bullying for alternatives such as reading in or severing some provisions to be workable. The declaration of unconstitutionality is to take effect immediately, because “[t]o temporarily suspend [it] would be to condone further infringements of Charter-protected rights and freedoms.” [220] Besides, the victims of cyber-bullying still “have the usual—albeit imperfect—civil and criminal avenues available to them.” [220]

I believe that this is the right outcome. However, Justice McDougall’s reasons are not altogether satisfactory. More on that soon.