Category: Law of Democracy

The Made-Up Law Made Them Do It

The Supreme Court’s made-up right to vote doctrine works its mischief at the Ontario Court of Appeal

Earlier this week, the Court of Appeal for Ontario released its decision in Working Families Coalition (Canada) Inc v Ontario (Attorney General), 2023 ONCA 139, which considers the constitutionality of an extension, from six months before an election to a whole year, of the period during which political speech by civil society actors in Ontario is severely restricted. The Superior Court had previously found that this extension was an unconstitutional violation of the freedom of expression protected by s 2(b) of the Canadian Charter of Rights and Freedoms. However, the Ontario legislature re-enacted it, invoking the Charter’s “notwithstanding clause”, s 33. The Court of Appeal unanimously holds that s 33 was validly relied on, but also, by a 2-1 majority, that the law nonetheless violates the right to vote, protected by s 3 of the Charter, whose application cannot be ousted under s 33.

The outcome is a disturbing one. The idea that a law that does not affect anyone’s ability to cast a ballot or run for office ― the two rights protected by s 3 ― but rather censors individuals and groups who are not candidates at an election precisely because they are not candidates, is a violation of the right to vote is, to put it mildly, counter-intuitive. The problem with the impugned legislation is that it is rank political censorship. Yet one would think that, since it enables legislatures to disregard the freedom of expression, s 33 of the Charter enables just this sort of self-serving abuse of power. Yet it would be a mistake to blame the Court of Appeal. The majority’s decision is a plausible application of one of the Supreme Court’s worst decisions of the last half-century: Harper v Canada (Attorney-General), 2004 SCC 33, [2004] 1 SCR 827.

Before getting to the main issue, a few words on the unambiguously correct and good aspect of the Court of Appeal’s ruling: the rejection of the argument that s 33 could not have been invoked in the first place. This too is a straightforward application of Supreme Court precedent, Ford v Quebec (Attorney General), [1988] 2 SCR 712, which held that the only implicit limit on resort to s 33 is that it cannot be retroactive; subject to this constraint, legislatures need not explain or justify suspending the enforceability of Charter rights. The applicants argued that Ford could be disregarded, either because election law was a special case or because the Supreme Court’s decision was no longer in tune with “the evolution of Charter jurisprudence since … 1988″ [56]. The Court of Appeal makes short work of both arguments, explaining that the importance of elections to the maintenance of democracy is sufficiently addressed by the fact that s 3 of the Charter is not subject to s 33, and that Ford has never been questioned, let alone overruled, by the Supreme Court.

This is quite right on both points. The fashionable academic theories on which the applicants relied, developed in the last few years in response to the resurgence of s 33, are unmoored from the Charter‘s text, and rely on fanciful extension of underlying principles about whose effects the Supreme Court is ambivalent at best. Of course, the Supreme Court remains free to make things up and reverse Ford. It may yet be urged to do so, whether if this case is appealed or indeed, as some have suggested, in a reference intended to limit the use of s 33. But I hope that the ease with which the Court of Appeal rejected the claim that Ford has been superseded by jurisprudential developments is indication of what is to come if that happens.

Section 2(b) of the Charter having been successfully ousted, the Court of Appeal moves on to the main event: the s 3 argument. This too is governed by Supreme Court precedent, Harper, which concerned the constitutionality of the federal scheme for silencing civil society political speech during (but not prior to) election campaigns. But the guidance it provides is nothing as clear as Ford‘s, and it is necessary to reproduce it here at some length.

Harper was mainly argued and decided on the basis of s 2(b), but s 3 was also raised. Bastrache J’s majority reasons on this point began by noting that it “cannot be” that “the right to meaningful participation” in elections, which is how the Supreme Court has long re-interpreted s 3, has an identical content “with the exercise of freedom of expression. … The right to free expression and the right to vote are distinct rights”. [67] Would that Justice Bastarache had stopped here! Instead, he declared that “[t]he right to meaningful participation includes a citizen’s right to exercise his or her vote in an informed manner”. [70] This drew on no s 3 precedent whatever, but rather on Libman v Quebec (Attorney General), [1997] 3 SCR 569, a s 2(b) case. Undeterred, Ba starache J had the following to say:

[E]quality in the political discourse is necessary for meaningful participation in the electoral process and ultimately enhances the right to vote. Therefore … s. 3 does not guarantee a right to unlimited information or to unlimited participation. Spending limits, however, must be carefully tailored to ensure that candidates, political parties and third parties are able to convey their information to voters. Spending limits which are overly restrictive may undermine the informational component of the right to vote. To constitute an infringement of the right to vote, these spending limits would have to restrict information in such a way as to undermine the right of citizens to meaningfully participate in the political process and to be effectively represented.

The question, then, is whether the spending limits … interfere with the right of each citizen to play a meaningful role in the electoral process. The trial judge found that the advertising expense limits allow third parties to engage in “modest, national, informational campaigns” as well as  “reasonable electoral district informational campaigns” but would prevent third parties from engaging in an “effective persuasive campaign” … [But] [m]eaningful participation in elections is not synonymous with the ability to mount a media campaign capable of determining the outcome. [72-74; paragraph break removed]

The outcome of Working Families turns on the meaning of this less-than-pellucid passage, of which the majority and the dissent take different views.

The majority, Zarnett and Sossin JJA sees it as setting out “two proxies, or methods of ascertaining whether the restriction” on voter information “is constitutionally offside”. [86] The first is asking whether the restriction is, in Bastarache J’s words, “carefully tailord”, which in turn “”The requirement that the restriction be carefully tailored “invites the court to examine the rationale, express or implicit, for the amount and duration of the spending limit – the express or implicit reasons why the lines were drawn where they were”. [87] This, the majority insists, is a very different matter from the analysis required by s 1 of the Charter, and in particular from its “minimal impairment” stage, which asks whether less restrictive alternatives to the impugned measure were (reasonably) available to the legislature. Here,

the question is whether the challenged spending restrictions draw the line at the point of preventing the well-resourced from dominating political discussion without being overly restrictive so as to undermine the right of citizens to meaningfully participate in the political process and to be effectively represented. A conclusion that a choice was in some other sense “reasonable” does not answer this question. [89]

Moreover, in this case, this assessment must focus not only on the legal end-state created by the impugned law, but specifically on the transition from the regulatory regime it displaced, which restricted political speech by civil society actors for six months rather than a year. It is the change from the one to the other that must “carefully tailored” in the above sense.

The majority holds that it was not. While explicitly rejecting the contention that the impugned legislation “constitutes partisan self-dealing by the incumbent government”, [102] it considers that “doubling the restricted period without increasing the quantum, a result that was twice as restrictive as what had been found appropriate, without explanation, does not denote careful tailoring”. [109] While the government argued that the new regime did not impede the voters’ participation, the majority takes the position that “[i]f at least some voters are prevented from exposure to political information of value from third parties in the 6 to 12-month period, their right to meaningful participation under s. 3 may be undermined”. [112] That the new restriction is one of a range of reasonable alternatives does not matter either ― that would be a consideration under s 1 of the Charter, but not at the point of establishing a s 3 infringement.

The second “proxy” is whether the restrictions leave room for at least “a modest informational campaign”. The majority finds that there was no evidence that this was so. The first-instance judge’s suggestion that affordable means of communicating with the voters were available and sufficient for a modest campaign was speculative. Moreover, the resources that could be used under the impugned law had to be deployed over a period of 12 months, which again threw the validity of the law into doubt.

Having briefly considered whether the restriction of s 3 rights could be justified under s 1 of the Charter, the majority concludes that it could not. The law is unconstitutional, but the declaration to this effect is suspended for a year to allow the legislature time to consider its next steps.

In dissent, Benotto JA rejects the majority’s interpretation of Bastarache J’s reasons in Harper. For him

[t]he controlling test is not whether the spending limits are carefully tailored but whether they restrict information in such a way to undermine the right of citizens to meaningfully participate in the electoral process, which includes the right to vote in an informed manner. [161]

This test is concerned with the effects of the impugned law, not with whether a justification for it exists. To look at justification is to conflate the s 3 analysis with that which ought to take place under s 1. In this case, moreover, it would be a mistake to focus on the change from s six-month period of restricting civil society speech to the one-year one; the longer period “had to stand or fall on its own. It was not the change that was determinative, but whether the legislation … was Charter compliant.” [176]

The dissenting judge considers there was enough evidence that the impugned law left some space for civil society actors to communicate their views to the voters, which was all that Harper required. The judge below made findings to this effect which were open to him and should not be disturbed.

I have no strong views on whether the majority opinion or the dissent is the better application of Bastarache J’s comments in Harper. I think both the majority’s reading, which emphasises the importance of the “careful tailoring” language and the dissent’s, which focuses on the way Bastarache J seems to have formulated the ultimate question before him are plausible. It is true, as the dissent charges, that the majority’s “careful tailoring” analysis is hard to tell apart from what would normally take place under s 1 of the Charter. I would add, moreover, that the “two proxies” approach will be unhelpful if the two point in different directions, which one might think was the case here: the law wasn’t tailored carefully, or indeed at all, but it arguably did leave some room for political speech. But the approach favoured by the dissent suffers from its own flaws. For one thing, it seems to ignore Bastarache J’s tailoring language altogether. For another, it is entirely impressionistic, and leaves an ostensible constitutional right at the mercy of the government producing an expert who will say, as a former Chief Electoral Officer did in this case, that the spending limit imposed on civil society was “not nothing”. Pick your poison.

For my part, I want to stress that this case highlights the rare feat achieved by Bastarache J (and, of course, the other judges who signed onto his opinion) in Harper: being at once vapid and pernicious. Vapid, because the discussion of s 3 in Harper is too vague and self-contradictory to mean much of anything, let alone provide real guidance to the courts that are nonetheless bound to apply it. To repeat, it is not the Court of Appeal judges’ fault that they have a hard time puzzling out whether “careful tailoring”, “modest informational campaign”, or “meaningful participation” is the test for a s 3 violation, and what any of these things mean. Pernicious, because it still opens the door to what is quite obviously a freedom of expression issue that should be dealt with under s 2(b) of the Charter ― or, as here, ignored because the self-dealing legislature so decreed ― to be considered under the aegis of a different right, unsuited to the exercise as a matter of both constitutional text and doctrine.

Of course it’s true that, as Bastarache J said in Harper, “[g]reater participation in the political discourse leads to a wider expression of beliefs and opinions and results in an enriched political debate, thereby enhancing the quality of Canada’s democracy”. [70] But it simply does not follow that this is a matter for s 3 of the Charter and that “the right to vote in an election of members of the House of Commons or of a legislative assembly” includes a “right to exercise his or her vote in an informed manner”. [71] Not everything that is needed to make a given Charter right fully effective can be rolled into that particular right. A free press, and certainly the media’s ability to report on court proceedings, “enhance the quality” of the administration of justice and, for instance, the right to be judged by an independent and impartial tribunal. But it does not follow that restrictions on reporting on criminal trials are to be dealt with under s 11(d) of the Charter instead of s 2(b). Different Charter provisions have independent meanings and distinct doctrinal frameworks that give them effect, and confusing them is both wrong in principle and unhelpful in practice ― except, of course, for crassly results-oriented purposes.

In another controversy about election laws in Ontario, the Supreme Court put an end to similar confusion. In Toronto (City) v Ontario (Attorney General), 2021 SCC 34, it rejected the Superior Court’s re-branding of the franchise in municipal elections, to which s 3 of the Charter does not apply, as a form of expression protected by s 2(b). If given the opportunity, it should do the same with the re-branding of pre-electoral expression as “the right to vote in an election of members … of a legislative assembly”. This should be done in the clearest way possible ― that is to say, by rejecting Harper, at least on this point (until, in the fullness of time, its s 2(b) holding is also overturned). Harper‘s s 3 “analysis” was made-up, and it needs to be unmade in the place whence it came.

If It’s Broke, You’re Not the One to Fix It

The Québec Court of Appeal takes it upon itself to update obsolete election legislation. That’s not its job.

This post is co-written with Mark Mancini

One of us (Sirota) has written any number of times about Québec’s Election Act, which is remarkable by the staggering restrictions it imposes on election campaigns and by its drafting that has, on many points, not been updated this century. This combination of severity and obsolescence leads to all manner of controversy and problems in the Act’s application. A recent decision of the Québec Court of Appeal, Therrien c Directeur général des élections du Québec, 2022 QCCA 1070, illustrates this. 

At issue in Therrien was s 429 of the Act, which provides that, in the week after the writ for an election

is issued, no person, except the Chief Electoral Officer [CEO], may broadcast or cause to be broadcast by a radio or television station or by a cable distribution enterprise, publish or cause to be published in a newspaper or other periodical, or post or cause to be posted in a space leased for that purpose, publicity relating to the election.

As the Court (Justice Cournoyer writing with the agreement of Justice Dutil; Justice Fournier, who had also been on the panel, passed away before the decision was issued) recognizes, “when s 429 … was amended in 1995, social media did not exist. … The ordinary meaning of the words ‘post’ and ‘space leased’ could not envision virtual reality, be it virtual posters or virtual spaces”. [62]-[63] (We translate here and throughout.) The question was whether s 429 nonetheless applied to prohibit advertising on Facebook, such as an ad that the CAQ, for which the appellant was the social media manager, took out in the first week of the 2014 election campaign.

2014, you might think, is a long time ago. We’ll return to this below. You might also think that s 429 is unconstitutional. We are inclined to think so too. In Thomson Newspapers Co v Canada (Attorney General), [1998] 1 SCR 877, the Supreme Court struck down a publication ban on polls for part of an election campaign, and a part both more sensitive and shorter than the one at issue here, namely the last three days. It is hard to argue that a ban on some advertising in the campaign’s first week is more justified, although perhaps a court would accept that it is necessary to maintain fair electoral competition. But the issue does not seem to have been raised in Therrien, which is a pure case of statutory interpretation. The Court observes that the issue of the applicability of a provision to circumstances that were not and could not have been anticipated at the time of its enactment is “a classic in statutory interpretation, but its solution, as this case shows, is not always obvious”. [9] With this much we agree. The Court’s solution in this case is that s 429 does apply to social media advertising. This we believe is wrong, in light of the―to repeat, obsolete―drafting of the Act.

The Court begins by interpreting s 429 on its own terms. Its effect is to ban some―though not all―political advertising in the first week of an election campaign. Its purpose, inferred from what it prohibits, is “to foster fairness and equality among all political parties at the outset of an election campaign”, [54] by preventing the incumbent from getting a jump-start on its competitors. As the Court notes, fixed election dates weren’t in place when s 429 was enacted.

Inferring this purpose from the mischief sought to be remedied is an unremarkable tool of interpretation, but in this case, we fear the Court’s analysis is backwards. It may be true that the purpose of the provision is to foster fairness and equality at the outset of the campaign. But this purposive analysis must build on convincing evidence in the text and the choices reflected in that text. In this sense, the Court’s analysis is flipped. At a number of points, it puts no weight on the ordinary, accepted meaning of the text, seemingly allowing the Court’s own view of the statutory purpose to drive the analysis. 

This leads the Court down the incorrect path. Drawing on Perka v R, [1984] 2 SCR 232 and R v 974649 Ontario Inc, 2001 SCC 81, [2001] 3 SCR 575, it states that while a statute’s terms are to be given the meaning they had at the time of their enactment, they “must not necessarily be confined to their original meaning” [65] at that time. What this means is that statutory language, provided it is sufficiently general, can be applied to facts and phenomena that weren’t or couldn’t be contemplated when it was enacted. But the focus of the analysis must be on the language used in the statute, and whether it could conceivably cover the phenomenon at issue. The issue, then, is “whether the text of s 429 prevents its extent to virtual posting in a virtual space”. [67]

The Court is right to cite these authorities at the outset, for they confirm the basic rule: the original meaning of statutory terms governs. This point was expressed most recently in R v Kirkpatrick, 2022 SCC 33, in the concurring opinion of Coté, Brown, and Rowe JJ. The concurrence articulated the accepted rule, unchallenged by the majority:  “[i]t is a fundamental error to apply the ‘living tree’ methodology to the interpretation of statutes” [55]. But the Court of Appeal disregards the basic principle it cites. Rather than asking whether the words can bear the “adaptation by the courts of general concepts to these new realities” [68], it expressly concludes that the meaning of the words “post” and “space leased” “could not contemplate virtual reality” [63]. It then moves to conclude that the terms “post” and “space” “do not prevent their application to the virtual dimension specific to social networks” [67].

Here both the method and the conclusion are faulty. As we note, the accepted method asks whether the provision, in its purposive context, can accommodate the new technological developments. The Court, instead, reasons backwards: instead of asking “does this provision apply?”, it asks “why wouldn’t this provision apply?”. This is inappropriate on several grounds.  Most basically, it is always for the party alleging that a provision applies―here, the CEO―to prove that this is so, and for good reason. Legislators who vote for legislative proposals do not and cannot time travel. The reach of statutes is fundamentally limited by their wording. By failing to positively affirm that a provision applies in a given circumstance, the Court distorts the reach of the law to cover phenomena that the text simply may not support. This, as we shall see, is an unacceptable form of spurious interpretation.

There are other normative reasons to reject the Court’s interpretation. Since the provision at issue is a penal one and restricts political speech, both the rule of lenity and the principle of legality counsel against applying it to doubtful or borderline cases. And substantively, the idea of a “virtual space” isn’t just a novel application of the existing concept of a space in the way that, say, same-sex marriage is a new application of the old concept of marriage. It is a metaphor and cannot do the work the Court wants it to.

The Court’s so-called purposive approach is also left wanting on its own terms, as it fails to have proper regard to the legislative context and to show why the purpose of the provision compelled its chosen interpretation. Consider the Court’s analysis of the history of s 429. In Frank v Canada, 2019 SCC 1, [2019] SCR 3, the dissenting opinion (unchallenged on this point), noted that “[t]he state of the law as it existed prior to an impugned provision coming into force can…give insight into why the provision was enacted” [131]. This, of course, is one way to discern the meaning of text; changes in wording can indicate changes in legislative purposes (as opposed to inferences based on what a legislature did not do).  In this case, the appellant sought to draw the Court’s attention to the fact that s 429’s predecessor provisions were phrased in general terms and did not specify particular forms of advertising prohibited in the early campaign, arguing that the legislature’s choice to now ban some forms of ads and not others had to be respected. The Court simply isn’t interested: “the history of the amendments to s 429 does not matter as much as the parties think in interpreting its text”. [55] In our view, this is a mistake. As noted in the Frank dissent, the history of a provision can often illuminate the textual means by which a legislature was attempting to solve a particular mischief. If the appellant is right (and the Court does not even bother setting out the previous version of s 429, so we cannot tell), his argument deserved to be taken seriously.

The Court goes on to add that its interpretation of s 429 is in agreement with that of the CEO, which can be taken into account without being binding. It is a bit difficult to say how much this argument influenced the Court―it is probably not a major factor in the decision. But any reliance on it is, nonetheless, disturbing. A court would not take special notice of the police’s interpretation of the Criminal Code. There is no reason to treat an administrative enforcement agency with any more indulgence. (It is telling, too, that the case on which the Court relies here, Cayouette c Boulianne, 2014 QCCA 863, is at root a dispute among neighbours, which turns on the meaning of a municipal by-law. Giving some weight to the municipality’s views in that context is not nearly as problematic as doing that when the administrator is the prosecutor.)

All in all, the Court’s analysis on this point is backwards as a matter of method, but the result is also problematic. Some may ask why the original meaning rule should be followed in a case like this, where new technological problems are so evident. The answer relates to the point of statutory interpretation. The job of courts is to interpret the text through which legislatures seek particular objectives (MediaQMI v Kamel, 2021 SCC 23, [39]). The text discloses how a legislature wanted to achieve its ends. By updating the statute for the legislature, the Court assumes that the legislature (a) wants its law extended; and (b) wants the law extended in this particular manner. It deprives the legislature—the exclusive law-maker—of the opportunity of creating a new regime that balances on- and offline expression. Citizens can rightly begin to question where the law is made.

The Court also accepts an alternative argument based on the effect of Québec’s Act to establish a legal framework for information technology (IT Framework Act) on s 429. In a nutshell, this law is meant to ensure that digital documents are treated the same as their analogue counterparts for various purposes. Documents are defined as follows, in s 3 of the IT Framework Act:

Information inscribed on a medium constitutes a document. The information is delimited and structured, according to the medium used, by tangible or logical features and is intelligible in the form of words, sounds or images. The information may be rendered using any type of writing, including a system of symbols that may be transcribed into words, sounds or images or another system of symbols.

Moreover, pursuant to s 71

The concept of document, as used in this Act, is applicable to all documents referred to in legislative texts whether by the term “document” or by terms such as act, deed, record, annals, schedule, directory, order, order in council, ticket, directory, licence, bulletin, notebook, map, catalogue, certificate, charter, cheque, statement of offence, decree, leaflet, drawing, diagram, writing, electrocardiogram, audio, video or electronic recording, bill, sheet, film, form, graph, guide, illustration, printed matter, newspaper, book, booklet, computer program, manuscript, model, microfiche, microfilm, note, notice, pamphlet, parchment, papers, photograph, minute, program, prospectus, report, offence report, manual and debt security or title of indebtedness.

The Court holds that

the concept of document necessarily includes virtual posts, because the posts consist of information inscribed on a medium which has the same legal significance if it includes the same information, regardless of the medium … In this respect, “to post” or “cause to be post” includes the use of a medium on which information is inscribed, i.e. a document within the meaning of s 3.  Meanwhile … the absence of words “poster”, “post”, or “cause to post” from the list in s 71 is of no consequence. The use of the phrase “such as” to introduce the list of many types of document is clearly aimed at excluding any restrictive interpretation of the term document, as defined in s 3. [86]-[87] (Paragraph break removed)

Here too we are not persuaded. For one thing, open-ended though it may be, we do not think that the IT Framework Act’s definition of a document extends to virtual posts, or any other media of a broadcast nature. The IT Framework Act’s purpose provision, s 1, refers to “documentary communications between persons, associations, partnerships and the State”. Elsewhere, the IT Framework Act speaks of documents producing “legal effect” or having “legal value” (e.g. ss 5 and 9). A poster―or a radio or TV ad―aren’t “documents” within the IT Framework Act’s meaning any more than in ordinary language.

Section 71 supports this view, although of course the Court is right that its enumeration is not strictly speaking closed. It is, however, remarkably exhaustive (which is why we thought it worthwhile to reproduce it above). And, tellingly, while it does include audio and video “recordings”, it does not include broadcasts. Considering the exhaustiveness, the fastidiousness even, of the enumeration, we do not think the omission is accidental or that it can be gotten around by relying on the “such as” language. This might not be the proverbial elephant, but we do not think the National Assembly hid a beaver in s 71’s mousehole. At minimum, the Court had to explain in what sense a virtual post is “such as” the objects enumerated in s 71, and it does not do this.

So much for Therrien itself. But we think it is important to point out that it is not an isolated case, but rather part of a pattern of very questionable decision-making by both the Québec Court of Appeal and successive CEOs with respect to the Act, which is in dire need of reform. In effect, those in charge of administering the Act are trying to maintain or even extend its reach while avoiding, on a case-by-case basis, consequences they find intolerable.

So far as the Court of Appeal is concerned, we have in mind the decision in Métallurgistes unis d’Amérique (FTQ), section locale 7649 c Québec (Directeur général des élections), 2011 QCCA 1043, which upheld the Act’s draconian restrictions on “third party” political spending. In that case, a union was fined for criticizing a political party in communications addressed to its own members. More generally, individuals who are not candidates and unincorporated groups are limited to spending 300$ on election advertising. Corporations, including not-for-profit ones, are prohibited from spending a penny. This is difficult to reconcile with the Supreme Court’s decisions in Libman v Quebec (Attorney General), [1997] 3 SCR 569 and Harper v Canada (Attorney General), [2004] 1 SCR 827, which recognized the right of “third parties” to engage in electoral advertising even as they also accepted the principle that such advertising can be strictly limited.

As for the CEOs, they attempted to censor “third party” interventions in each of the last two election campaigns ― that of a group opposed to the then-proposed “Charter of Values” in 2014 and that of environmentalist NGO Équiterre in 2018 ― provoking a public outcry. In 2014, the then-CEO flip-flopped and ended up withdrawing his objections. The 2018 story has only concluded recently (egregious delay is also, it seems, a pattern with the CEO), as Laura Lévesque and Thomas Laberge report in Le Soleil. The CEO has “blamed” and warned Équiterre but apparently not fined it. Yet as Sirota wrote at the time, the CEO was right to find fault with the campaigners on both occasions.

As further discussed in this post on the 2014 climb-down, the then-CEO reinterpreted the relevant provisions in a way that may have been sensible in light of the social and technological change since their enactment, as well as protective of the freedom of expression, but was not tenable in light of their text. The choice of merely “blaming” Équiterre is also, at first glance, understandable on the merits but not something provided for by the Act, except presumably as an exercise of an implicit prosecutorial discretion. In effect, the CEO is deciding what the Act means and when―and his decision to go easy on fairly clear violations by NGOs while prosecuting a debatable one by a political party is worth highlighting.

All this suggests, unequivocally to our mind, that the Act needs to be reformed so as to accommodate the social and technological realities of the 21st century. As it happens, the Canadian Press’s Jocelyne Richer reports that the CEO wants the Act to be “updated”―but mainly so as to introduce even more restrictions, specifically on advertising during the “pre-campaign” period. (In fairness, he was already asking for such an “update” in 2016. So far, the National Assembly has not obliged.) Parliament has added such restrictions to the Canada Elections Act some years ago, and Ontario has used the Charter’s “notwithstanding clause” to extend its censorship regime, which now covers more than one year in every four.

These rules are bad and possibly unconstitutional, as Sirota argued here and here. But, quite apart from their other problems, they would hit especially hard in Québec unless the Electoral Act’s existing strictures are relaxed to some degree, and also unless it is re-drafted so as to be technologically neutral to the extent possible. In the meantime, however, it is not the role of either the CEO himself or the courts to fiddle with the Act to make it work better. The law is broke, but they are not the ones who have the tools to fix it.

Correct, but Wrong

Thoughts on the Supreme Court’s dismissal of the challenge to Ontario’s interference in the Toronto municipal elections

Last week, the Supreme Court handed down its judgment in Toronto (City) v Ontario (Attorney General), 2021 SCC 34, upholding a provincial statute cutting the number of wards and councillors in Toronto in the middle of a municipal election campaign. The Court divided 5-4, with Chief Justice Wagner and Justice Brown writing for the majority (also Justices Moldaver, Côté, and Rowe) and Justice Abella for the dissent (also Justices Karakatsanis, Martin, and Kasirer).

The majority gets the outcome right. As both co-blogger Mark Mancini and I have written here in response to the Superior Court’s decision in this case, the province was well within its rights to enact what was, by all accounts, a disruptive law of questionable usefulness. But the majority’s reasoning is underwhelming. It’s not bad on the first issue: that of an alleged violation of the freedom of expression. But it is just rubbish on the second: that of the constitutional principle of democracy. The majority’s attempt to synthesize and cabin the Supreme Court’s jurisprudence on unwritten principles is a complete failure.

The first issue is whether the reorganization of the Toronto city council after the beginning ― though well over two months before the end ― of the municipal election campaign was a limitation of the freedom of expression of the candidates who had started campaigning under the old system. The majority holds that it was not. As a result, it does not get to the question of whether a limitation would have been justified.

For the majority, the matter falls to be considered as a claim for the provision by the state of a particular platform for expression, rather than as a (more usual) claim that a person is being prevented from conveying their ideas to others. As the majority explains,

the City does not seek protection of electoral participants’ expression from restrictions tied to content or meaning … ; rather, it seeks a particular platform (being whatever council structure existed at the outset of the campaign) by which to channel, and around which to structure, that expression. [32]

In other words, this is a “positive” rather than a “negative” right claim. The majority reformulates the test for such a claim as whether it is

grounded in the fundamental Charter freedom of expression, such that, by denying access to a statutory platform or by otherwise failing to act, the government has either substantially interfered with freedom of expression, or had the purpose of interfering with freedom of expression? [25]

The majority adds that “substantial interference with freedom of expression requires “effective preclusion” of “meaningful expression”, which is “an exceedingly high bar that would be met only in extreme and rare cases”. [27] 

The City has not cleared this bar. The majority states that “the candidates and their supporters had 69 days — longer than most federal and provincial election campaigns — to re‑orient their messages and freely express themselves according to the new ward structure”, with “no restrictions on the content or meaning of the messages that participants could convey”. [37] There was a meaningful election campaign, albeit a different one than had originally been planned.

The majority also rejects the City’s alternative argument on freedom of expression, to the effect that it implies a guarantee of “effective representation” which the Supreme Court originally articulated in the context of section 3 of the Charter. This provision protects the right to vote in federal and provincial ― not municipal ― elections. For the majority, “[e]ffective representation is not a principle of s. 2(b), nor can the concept be imported wholesale from a different Charter right”.

The dissent, for its part, begins by stressing the disruptiveness of the reform imposed by the province, and the lack of justification for it ― indeed, the new electoral structure had been considered by the City itself, and rejected. It goes on to argue that

When a democratic election takes place in Canada, including a municipal election, freedom of expression protects the rights of candidates and voters to meaningfully express their views and engage in reciprocal political discourse on the path to voting day. … When the state enacts legislation that has the effect of destabilizing the opportunity for meaningful reciprocal discourse, it is enacting legislation that interferes with the Constitution. [115]

This is what the province has done here, as the dissent emphasizes by quoting at great length the statements of candidates impacted by the disruption.

The dissent also argues strenuously that the majority is wrong to see the dispute as being about the positive provision of a platform for expression, and so to apply a higher threshold of seriousness to the question of whether the freedom of expression has been infringed. Indeed, in its view

There is no reason to superimpose onto our constitutional structure the additional hurdle of dividing rights into positive and negative ones for analytic purposes. Dividing the rights “baby” in half is not Solomonic wisdom, it is a jurisprudential sleight-of-hand that promotes confusion rather than rights protection. [155]

The province has also failed to advance a justification. This means that the impugned law contravenes the Charter.

As noted above, I think that the majority is basically right, notably in treating the claim advanced by the city as being for the provision or maintenance of a specific set of arrangements within which expression is to be channelled. The freedom of expression is the ability to say things one thinks, and not to say things one doesn’t. It’s not a guarantee that what one says will be interesting or relevant to anyone. If a province goes dry tomorrow, a great deal of alcohol advertising will have been rendered pointless, as will a great deal of campaigning for moderate drinking, research into the health benefits of red wine, and what not. But prohibition will not infringe the Charter. (It will be abominable, but constitutional.) It is the same when a province renders pointless a great deal of campaigning for a municipal election. Stupid, but constitutional, as Justice Scalia used to say.

The dissent’s response to this would be, I think, that the context of an election is different, but that really just proves the majority’s point. The claim at issue is about a specific platform for expression. The dissent’s analogy with Greater Vancouver Transportation Authority v Canadian Federation of Students — British Columbia Component, 2009 SCC 31, [2009] 2 SCR 295 also doesn’t work. That case was concerned with a ban on political advertisements on city buses, and the issue, as the majority explained, was not access to the platform in question ― that is, advertising on buses ― but a restriction on the content of what could be said on that platform. Here, the situation is exactly the opposite. The province hasn’t changed how it regulates the content of municipal election campaigns, but instead has shut down the old platform for expression and substituted for it a different one.

My objection to the Chief Justice’s and Justice Brown’s reasons has to do not with what they do, but with some of the things they say. They describe the threshold at which the “positive” freedom of expression is engaged as “an exceedingly high bar that would be met only in extreme and rare cases”. This may be tantamount to reading this aspect of the freedom out of the doctrine entirely ― but they also say that it has, in fact, some value. This language of “extreme and rare cases” isn’t necessary here, and I don’t think it provides useful guidance for the future; the words are too imprecise and subjective. The other troubling aspect of the majority’s reasons is its mention ― seemingly in passing, but I suspect that it is with at least a measure of approval ― of the fact that the Charter‘s guarantee of freedom of expression “has been interpreted so broadly that the framework has been criticized for setting too low a bar for establishing a … limitation”. [16] This has nothing to with this case, since that broad framework traditionally traced to Irwin Toy Ltd v Quebec (Attorney General), [1989] 1 SCR 927, doesn’t apply. The Supreme Court is already far too accepting of limitations on the freedom of expression, and its possible willingness to restrict the freedom’s scope bodes ill.

I turn now to the second issue, that of whether interference with an ongoing municipal election is an unconstitutional violation of the democratic principle. This principle, which the City suggested required the provision of “effective representation” in the municipal context, as well as in the cases governed by section 3 of the Charter, would serve as a limit on the legislature’s ability to enact laws in relation to “Municipal Institutions in the Province” pursuant to section 92(8) of the Constitution Act, 1867.

The majority is unimpressed. It says that unwritten principles such as democracy “are … part of the law of our Constitution, in the sense that they form part of the context and backdrop to the Constitution’s written terms”. [50] However,

because they are unwritten, their “full legal force” is realized not in supplementing the written text of our Constitution as “provisions of the Constitution” with which no law may be inconsistent and remain of “force or effect” under s. 52(1) of the Constitution Act, 1982. Unwritten constitutional principles are not “provisions of the Constitution”. [54]

They can serve two functions: on the one hand, they can be used as aids in interpreting constitutional text; on the other, they can fill textual gaps. What they cannot do, the majority says, is directly invalidate legislation. To hold otherwise would be to “trespass into legislative authority to amend the Constitution”, [58] and to make an end-run around section 1 and 33 of the Charter, which allow, respectively, reasonable limitation of rights and legislative override of some of them, including, relevantly for this case, the freedom of expression.

To support its claim that principles have only interpretive and suppletive effects, the majority reviews various cases that might suggest otherwise. Notably, it dismisses the dissent on the legal question in the Patriation Reference, Re: Resolution to amend the Constitution, [1981] 1 SCR 753, which the Supreme Court later unanimously endorsed in Reference re Secession of Quebec, [1998] 2 SCR 217, on the basis that “while the specific aspects of federalism at issue there may not have been found in the express terms of the Constitution, federalism is“. [52] As for the Provincial Judges Reference, Reference re Remuneration of Judges of the Provincial Court (PEI), [1997] 3 SCR 3, it stands, the majority says, for the proposition that “where the constitutional text is not itself sufficiently definitive or comprehensive to furnish the answer to a constitutional question, a court may use unwritten constitutional principles as interpretive aids” [65] and “to fill a gap where provincial courts dealing with non‑criminal matters were concerned”. [66]

In this case, “the unwritten constitutional principle of democracy cannot be used to narrow legislative competence” over municipalities, which is “plenary” and “unrestricted by any constitutional principle”. [80] Moreover,

The constitutional status of municipalities, and whether they ought to enjoy greater independence from the provinces, was a topic of debate during patriation … In the end, municipalities were not constitutionalized, either in amendments to the Constitution Act, 1867 or by reference in the democratic rights enshrined in the Charter. … Were the unwritten democratic principle applied to require all elections to conform to the requirements of s. 3 (including municipal elections, and not just elections to the House of Commons or provincial legislatures), the text of s. 3 would be rendered substantially irrelevant and redundant. [81]

The dissent, again, sees matters differently. It points out that unwritten principles have been recognized as binding both in Canada and in other “Parliamentary” [166] constitutional systems. (The dissent thus does not mention the United States.) It insists that

unwritten principles are our Constitution’s most basic normative commitments from which specific textual provisions derive. … Constitutional text emanates from underlying principles, but it will not always be exhaustive of those principles. In other words, the text is not exhaustive of our Constitution. [168]

The dissent rejects the majority’s insistence on the primacy of the text. Unwritten principles are just as important. It is they that “assist in developing an evolutionary understanding of the rights and freedoms guaranteed in our Constitution” ― that is, they “make[]” the constitutional living “tree grow”. [179] As for the majority’s argument based on section 52 of the Constitution Act, 1982, it “is a highly technical exegetical exercise designed to overturn our binding authority establishing that unwritten constitutional principles are a full constitutional partner with the text”. [183]

For the dissent, in “rare” cases “unwritten principles may be used to invalidate legislation” that “elides the reach of any express constitutional provision but is fundamentally at odds with our Constitution’s ‘internal architecture’ or ‘basic constitutional structure'”. [170] As the dissent sees things, this is what happened in the Provincial Judges Reference, as well as in Trial Lawyers Association of British Columbia v British Columbia (Attorney General), 2014 SCC 59, [2014] 3 SCR 31. That said, the dissent does not say anything about the application of the democratic principle in this case, which it has already resolved on the Charter argument.

By my lights, this is the judicial equivalent of a Leafs-Bruins game, which both ought to lose, but one has to win, just because. Now, I think that the majority’s conclusion is correct as a matter of both precedent and principle. As the Supreme Court held in British Columbia v Imperial Tobacco Canada Ltd, 2005 SCC 49, [2005] 2 SCR 473, it would be wrong to apply an unwritten principle so as to expand the scope of a Charter right so as to directly contradict clear text. Imperial Tobacco concerned the protection against retroactive legislation, which the Charter reserved to criminal law. Here we are dealing with the right to vote, and its love child “effective representation”, which the Charter reserves to federal and provincial, not municipal, elections. So far, so good. But only so far.

The rest of the majority’s analysis ― which, of course, is quite unnecessary, because the passage from paragraph 81 quoted above is enough to dispose of this issue ― rests on wholly untenable distinctions. The majority says that federalism is unlike the other constitutional principles ― indeed, that it is not a constitutional principle but part of the constitution’s “structure” ― because “federalism is” “found in the express terms of the Constitution”, notably the division of powers between the Dominion and the provinces. But the same is true of democracy and of the rule of law. We can point to some provisions, such as sections 1 and 3-5 of the Charter for democracy (as well, of course, as all the provisions having to do with the House of Commons in the Constitution Act, 1867), and sections 9-11 of the Charter and 133 of the Constitution Act, 1867 (again, this is not an exhaustive list) to say that these principles too are found in the express terms of the Constitution, and hence their other “specific aspects .. not found” in those express terms can nonetheless be judicially enforced.

Similarly, the majority’s distinction between alleged “gap-filling” in the Provincial Judges Reference “where provincial courts dealing with non‑criminal matters were concerned” and invalidating laws on the basis of unwritten principles is humbug. So far as these courts were concerned, the only reason the laws reducing their judges’ salaries were invalid was unwritten principle.

Last but not least, as Mark has noted, the majority doesn’t even begin to address  Canada (Minister of Citizenship and Immigration) v Vavilov, 2019 SCC 65, with its clear statement that

Where a court reviews the merits of an administrative decision … the standard of review it applies must reflect the legislature’s intent with respect to the role of the reviewing court, except where giving effect to that intent is precluded by the rule of law. [23; emphasis added]

In other words, Vavilov says that the Rule of Law principle does invalidate legislation to the extent that (though only to the extent that) it would require an incompatible standard of review.

The majority also says that “The unwritten constitutional principle of the honour of the Crown is sui generis” because, it seems, it “arises from the assertion of Crown sovereignty over pre‑existing Aboriginal societies … and from the unique relationship between the Crown and Indigenous peoples”. [62] But the other principles, such as federalism (a sine qua non for Canada’s existence) and democracy and the Rule of Law (1688 and all that), can also boast “unique” historical pedigrees.

The majority’s other arguments fare just as badly as its attempts at splitting hairs with a blunt axe. Applying principles to invalidate laws does not trespass into constitutional amendment if principles were already part of the constitution as enacted, in 1867 and in 1982. While some applications may inappropriately compromise section 33 of the Charter ― which is arguably one reason why Imperial Tobacco approach to cases to which the Charter already speaks is correct ― others will not. Principles are not reducible to expanded forms of Charter rights. Federalism is of course the obvious case in point. As for section 1 of the Charter, foreign precedents, such as the Australian jurisprudence on the implied freedom of political communication, suggest that something like a proportionality analysis can be combined with unwritten principles. Again, though, principles are not just a beefed-up Charter. Perhaps the best argument the majority advances is the one based on the word “provisions” in section 52(1) of the Constitution Act, 1982, but ― without endorsing the dissent’s rant about “technical exegetical exercises” ― I think that it is undermined by section 52(2)’s suggestion that “the Constitution of Canada” is not limited to textual sources, to say nothing of the Supreme Court’s jurisprudence to this effect.

A word, finally, on the dissent. It advocates not only for living constitutionalism, which as readers will know I think is a misbegotten interpretive approach, but also, more precisely, for what I have described as “constitutionalism from the cave“. This is the view that the constitution’s text is just a pale shadow of the true constitution, which judges alone can, over time, discover and impose. As much as I think the majority’s attempt to swat constitutional principles away is unsound as a matter of both doctrine and, sorry, principle, this is not a tenable alternative.

Here we are, then, at the last chapter of this unfortunate saga. It began with institutional vandalism by the Ontario legislature, and concludes with a Supreme Court decision that, despite narrowly reaching the right outcome, may yet do considerable damage of its own. The majority’s statements on freedom of expression are worrying, and its discussion of constitutional principles ― admittedly, a difficult subject (I have had more to say on it here) ― is almost entirely wrong-headed. The dissent, meanwhile, is largely unmoored from the law throughout. The judicial end is not better than the legislative beginning.

Disinformation by Omission

Additional thoughts on the futility of regulatory responses to mis- and disinformation

In my last post, I wrote about the Canadian Forces’ attempts to manipulate public opinion, including by means of disinformation, and about the dangers of regulations ostensibly meant to counteract disinformation. I briefly return to the issue of disinformation to highlight an excellent, if frightening, essay by David French in his newsletter for The Dispatch.

Mr. French writes about the alarming levels of polarization and mutual loathing by political partisans in the United States. He argues that this results from a “combination of malice and misinformation”, which mean “that voters hate or fear the opposing side in part because they have mistaken beliefs about their opponents. They think the divide is greater than it is.” Mr. French observes that many Americans are stuck in a vicious cycle:

Malice and disdain makes a person vulnerable to misinformation. Misinformation then builds more malice and disdain and enhances the commercial demand for, you guessed it, more misinformation. Rinse and repeat until entire media empires exist to supply that demand. 

And, crucially, Mr. French points out that misinformation does not just consist of “blunt, direct lying, which is rampant online”. It also includes “deception by omission (a news diet that consistently feeds a person with news only of the excesses of the other side) and by exaggeration and hyperbole”, which can be “in many ways more dangerous than outright lies”, because they cannot easily be countered with accurate information. (This is why the rhetorical practice of “nutpicking” ― pointing to the crazies on the opposite side, and claiming that they represent all those who might share something of their worldview ― is so effective. The nuts are real! They might even be somewhat prominent and influential, though not as much as the nutpicker suggests. Nutpicking isn’t lying. But it is deceptive, and destructive.) 

And yet, Mr. French cautions against regulatory responses to this crisis, serious though it is:

there is no policy fix for malice and misinformation. There is no five-point plan for national harmony. Popular policies … don’t unite us, and there are always differences and failures to help renew our rage. Instead, we are dealing with a spiritual and moral sickness. Malice and disdain are conditions of the soul. Misinformation and deception are sinful symptoms of fearful and/or hateful hearts. (Paragraph break removed)

I think this is tragically right, even though I do not share Mr. French’s deep Christian faith. Call it heart or mind instead of soul; speak of moral error, indeed of immorality instead of sin; this all is secondary, to my mind. The point is that the fault is not in our laws, but in ourselves. And this is why, in my last post, I wrote that the government

cannot be trusted with educating citizens and funding media in a way that would solve the problems of the “environment that has created the disinformation crisis”. The solution must come from the civil society, not from the state.

As I wrote long ago in the context of hate speech, the law ― at least so long as it remains relatively cabined and does not attempt comprehensive censorship ― cannot counteract the corrosive “messages … sent by sophisticated, intelligent people”, who are able to avoid crude hate propaganda, or outright lies. The hint, the understatement, the implication, the misdirection, the omission are their weapons, and the shield against it must be in our hearts and minds, not the statute book.

We often think of regulation as a sort of magic wand that can do whatever we need, provided we utter the right sort of spell when wielding it. This is, of course, an illusion, and entertaining it only distracts us from working on the most difficult subject of all: our selves.

Disinformation and Dystopia

Whose disinformation efforts should we really fear―and why we should also fear regulation to stop disinformation

Mis- and disinformation about matters of public concern is much in the news, and has been, on and off, for the last five years. First kindled by real and imagined interference in election campaigns, interest in the subject has flared up with the present plague. Yesterday’s developments, however, highlight the dangers of utterly wrongheaded responses to the issue, one that would will lead to a consolidation of government power and its use to silence critics and divergent voices.

First, we get a hair-raising report by David Pugliese in the Ottawa Citizen about the Canadian Armed Forces’ strong interest in, and attempts at, engaging in information operations targeting Canadians over the course of 2020. Without, it must be stressed, political approval, and seemingly to the eventual consternation of Jonathan Vance, the then-Chief of Defence Staff, the Canadian Joint Operations Command sought to embark on a “campaign … for ‘shaping’ and ‘exploiting’ information” about the pandemic. In their view “the information operations scheme was needed to head off civil disobedience … and to bolster government messages”. They also saw the whole business as a “learning opportunity” for what might become a “routine” part of their operations.

Nor is this all. At the same time, but separately, “Canadian Forces intelligence officers, culled information from public social media accounts in Ontario”, including (but seemingly not limited to) from people associated with Black Lives Matter. This, supposedly, was “to ensure the success of Operation Laser, the Canadian Forces mission to help out in long-term care homes hit by COVID-19 and to aid in the distribution of vaccines in some northern communities”. A similar but also, apparently, unrelated effort involved the public affairs branch of the Canadian Forces, which want its “officers to use propaganda” peddled by “friendly defence analysts and retired generals” and indeed “information warfare and influence tactics”, “to change attitudes and behaviours of Canadians as well as to collect and analyze information from public social media accounts” and “to criticize on social media those who raised questions about military spending and accountability.”

And in yet another separate incident,

military information operations staff forged a letter from the Nova Scotia government warning about wolves on the loose in a particular region of the province. The letter was inadvertently distributed to residents, prompting panicked calls to Nova Scotia officials … [T]he reservists conducting the operation lacked formal training and policies governing the use of propaganda techniques were not well understood by the soldiers.

To be blunt, there seems to be a large constituency in various branches of the Canadian forces for treating the citizens whom they are supposed to defend as enemies and targets in an information war. Granted, these people’s enthusiasm seems to outstrip their competence ― but we know about the ones who got caught. We can only hope that there aren’t others, who are better at what they do. And it’s not a happy place to be in, to be hoping that your country’s soldiers are incomptent. But here we are.

Also yesterday, as it happens, the CBA National Magazine published the first episode of a new podcast, Modern Law, in which its editor, Yves Faguy, interviewed Ève Gaumond, a researcher on AI and digital technologies, about various techniques of online persuasion, especially during election campaigns. These techniques include not only mis- and disinformation and “deep fakes”, but also advertising on social media, which need not to untruthful, though it may present other difficulties. Mr. Faguy’s questions focused on what (more) should Canada, and perhaps other countries, do about these things.

Ms. Gaumond’s views are somewhat nuanced. She acknowledges that “social media is not the main driver of disinformation and misinformation” ― traditional media still are ― and indeed that “we’re not facing a huge disinformation crisis” at all, at present. She points out that, in debates about mis- and disinformation, “the line between truth and falsehood is not so clearly defined”. And she repeatedly notes that there are constitutional limits to the regulation of speech ― for example, she suggests that a ban on microtargeting ads would be unconstitutional.

Ultimately, though, like many others who study these issues, Ms. Gaumond does call for more and more intrusive regulation. She claims, for instance, that “[i]f we are to go further to fight disinformation”, online advertising platforms should be forced not only to maintain a registry of the political ads they carry and of the amounts the advertisers spent, but also to record “[t]he number of times an ad has viewed” and “the audience targeted by the ad”. This would, Ms. Gaumond hopes, deter “problematic” targeting. She also wants to make advertising platforms responsible for ensuring that no foreign advertising makes its way into Canadian elections, and tentatively endorses Michael Pal’s suggestion that spending limits for online advertising should be much lower than for more conventional, and more expensive, formats.

Ultimatelty, though she doesn’t “think that we should tackle speech per se”, Ms. Gaumond muses that “[w]e should see how to regulate all platforms in a way that we can touch on all possible ways that disinformation is spread”. This means not only spending limits but also that “[y]ou cannot pay millions of dollars to microtarget … what you’re saying to people that believe the same thing as you do without oversight from other people, from Election Canada”. And beyond that

not only regulating social medias [sic], but also all of the environment that has created the disinformation crisis. That means education, funding and great journalism, the media ecosystem is one of the important components of why we’re not facing such a big disinformation crisis.

There are a few things to say about Ms. Gaumond’s proposals ― keeping in mind Mr. Pugliese’s report about the activities of the Canadian forces. The overarching point is the one suggested by the juxtaposition of the two: while researchers and politicians fret about disinformation campaigns carried ou by non-state and foreign actors, the state itself remains the most important source of spin, propaganda, and outright lies with which we have to contend. Unlike bots and Russian trolls, the state can easily dupe the opinion-forming segments of society, who are used to (mostly) believing it ― partly out of ideological sympathy, but partly, and it’s important to stress this, because the state is also an important source of necessary and true information which such people rely on and relay.

This means that we should be extremely wary of granting the state any power to control information we can transmit and receive. Its armed agents think nothing of manipulating us, including for the sake of propping up the government of the day. And it is no answer that we should grant these powers to independent, non-partisan bureaucracies. The Canadian Forces are also an independent, non-partisan bureaucracy of sorts. I’m pretty confident that they weren’t trying to manipulate opinion out of any special affection for the Liberal Party of Canada, say. They are just on the side of order and stability, and any civilian bureaucratic structure would be too. It would also be likely to be tempted to squish questions about its own budget and functioning, and to develop an unhealthy interest in people it regards as trouble-makers. Civilians might be more suspicious of right-wing groups than of BLM, but the ones have the same right to free speech and to privacy as the others.

Another thing to note is the confusion among the different issues clustered under the general heading of concerns about mis- and disinformation. Concerns about the targeting of advertising may be valid or not, but their validity often has little to do with the truthfulness of the ads at issue. Concerns about foreign influence may be magnified when it is being exercised through misleading and/or microtargeted ads, but they are not necessarily linked to the issues either of disinformation or of microtargeting. Spending limits, again, have little to do with disinformation. No doubt a knowledgeable researcher like Ms. Gaumond would be more careful about such distinctions in a paper than she sometimes is in the interview with Mr. Faguy. But can untutored policy-makers, let alone voters, keep track?

In light of all this, Ms. Gaumond’s suggestions, though sprinkled with well-intentioned caveats about “not saying ‘you cannot say that'”, should give us serious pause. Even increasing disclosure requirements is far from a straightforward proposition. As Ms. Gaumond notes, Google simply refused to carry political ads rather than set up the registry the government required. Facebook and Twitter might follow if they are forced to make disclosures that would reveal the functioning of their algorithms, which they may have good reasons for keeping out of their competitors’ sight. More fundamentally, the idea that all (political?) speech should at all times be tracked and monitored by the state does not strike me as healthy. Political debate is a fundamental right of citizens, not something we can only engage in on the government’s sufferance. We are not children, and government ― including Elections Canada ― is not a parent who needs to know what we are getting up to online. Last but not least, because of the government’s track record of spin and deceit, it cannot be trusted with educating citizens and funding media in a way that would solve the problems of the “environment that has created the disinformation crisis”. The solution must come from the civil society, not from the state.

Lastly, let me note in my view Ms. Gaumond may be far too optimistic about the willingness of Canadian courts to uphold constitutional limits on government regulation of electoral speech. Their record on this issue is generally abysmal, and the Supreme Court’s reasoning in the leading case, Harper v Canada (Attorney General), 2004 SCC 33, [2004] 1 SCR 827, is itself misinformed and speculative. If government actors take the initiative on these matters, the courts will not save us.

The issue of mis- and disinformation is at least much a moral panic as a genuine crisis. As Ms. Gaumond points out, the trouble is to a considerable extent with traditional media and political forces outside anyone’s easy control; as Mr. Pugliese’s reporting makes clear, we must fear our own government at least as much as any outside force. Yet fears of new technology ― not to mention fear-mongering by media and political actors whose self-interest suggests cutting social media down to size ― mean that all manner of new regulations are being proposed specifically for online political discussions. And the government, instead of being reined in, is likely to acquire significant new powers that will further erode the ability of citizens to be masters in their own public and private lives.

Common Power Grabs

A defence of Ontario’s use of the notwithstanding clause as “common good constitutionalism” is the same old tripe, under a new sauce

Over at Ius et Iustitium, Kerry Sun, Stéphane Sérafin, and Xavier Foccroulle Ménard (I shall refer to them collectively as SSM) have a new addition to the rather stale menu of notwithstanding clause apologetics: a post that attempts to justify legislative override of the Canadian Charter of Rights and Freedoms as a form of “common good constitutionalism”. SSM write

that the notwithstanding clause should be viewed as enshrining a form of coordinate interpretation. Under this approach, ideally, the invocation of s. 33 may be contemplated in those cases where a legislature seeking to advance the common good reasonably disagrees with the judicial interpretation of a rights provision

Except for the invocation of the “common good”, this is the usual fare. Legislatures are supposed to have their own views about what Charter rights mean and entail, and are justified in imposing these views on the citizens. Joanna Baron and Geoffrey Sigalet made one such argument over at Policy Options a couple of years ago (I critiqued it here), and more recently Professor Sigalet made a similar case in a National Post op-ed with Ben Woodfinden.

But the addition of the “common good constitutionalism” sauce is noteworthy. So far as it is possible to define, “common good constitutionalism” is a branch of right-wing anti-liberal thought which seeks to re-establish constitutional law on foundations ostentatiously grounded in traditionalist ideology and/or medieval natural law, and thereby to make it serve the general good, as understood by its exponents. In substance, “common good constitutionalism” often amounts to a celebration of political power at the expense of the rights of minorities. In form, it distinguishes itself not only by the aforementioned ostentatious traditionalism or medievalism, but also by its a refusal to seriously engage with non-adherents to the doctrine. (Its celebrity chef, Adrian Vermeule, is notorious for blocking people who have not attacked or sometimes even interacted with him on Twitter.)

Unfortunately, these traits are all present in SSM’s post. I address a number of specific faulty arguments it makes below, but first let me note that ― remarkably for a piece of scholarly writing ― it never quotes or even cites the people it disagrees with. They are merely nameless, faceless “critics” of this or that, and the only source SSM refer to for their views is the not-at-all critical op-ed by Professor Sigalet and Mr. Woodfinden linked to above. Mr. Ménard tries to make a virtue out of this in a subsequent Twitter exchange with Emmett Macfarlane, candidly admitting that he would “rather cite jurists who share” his fundamental premises “than political scientists with whom I share piecemeal views. It makes for better scholarship”, he says. No, it doesn’t. Participants in scholarly debate should endeavour to bring their opponents’ best arguments to their audience’s attention. Those who fail to do so risk becoming propagandists, no matter how many footnotes their writings include.

The entrée for SSM’s paean to the notwithstanding clause is the enactment by the Ontario legislature of the Protecting Elections and Defending Democracy Act, 2021, which invokes s. 33 of the Charter to override the decision of the province’s Superior Court of Justice in Working Families Ontario v Ontario, 2021 ONSC 4076. I will eventually post a detailed analysis of the Court’s decision, but as I have already noted in The Line, its conclusion is self-evidently correct. Section 1 of the Charter requires limits on the rights it guarantees to be reasonable and demonstrably justified. Yet the Ontario government simply provided no justification for extending the duration of very severe restrictions on the ability of civil society groups to engage in political advertising from six months before the start of an election campaign to a year. It own experts had previously said that the six-month period was reasonable. The law could not stand. But the legislature re-passed it in four days.

SSM’s presentation of the situation is misleading. For one thing, they claim that the “arguments” against Ontario’s legislation were “very similar to those raised in” Harper v Canada (Attorney General), 2004 SCC 33, [2004] 1 SCR 827. This is doubly wrong. First, the case actually decided by the Superior Court was focused on the lack of justification for the latest extension of Ontario’s censorship regime, not the validity of such a regime in principle. But even the original dispute about the six-month-long pre-campaign censorship period is far outside the scope of Harper. There the majority invoked the lack of restraint on political speech outside a five-week-long election campaign as evidence of the limited (and hence justified) nature of the restraints during that campaign. SSM further mislead their readers by suggesting that, “[a]s a result of the court ruling, Ontario would likely have no spending limits by unions, corporations, or other third parties in place prior to the next election period, slated to begin in the summer of 2022”. Needless to say, the Ontario legislature could have re-enacted a six-month (or shorter) restriction period just as easily is it re-enacted a year-long one. Its masters in the executive just chose not to do that.

This brings me to another weakness in SSM’s argument. Responding to critics of “the Ontario legislature’s failure to advance a justification for” invoking the “notwithstanding clause”, they insist that “a justification was in fact given in this case: preserving the fairness and integrity of Ontario’s provincial elections”. Leave aside its substantive merits for the moment, and notice the artful use of the passive voice: a justification “was advanced” ― by whom? The text does not say, but the footnote supporting this sentence refers to two sources. One is a passage from the Working Families judgment quoting the Attorney-General’s speech to the legislature about the bill it struck down; it simply has nothing to do with the use of the notwithstanding clause. The other is a news story quoting a statement by a spokesman for the government’s House Leader. Neither, in other words, reflect the legislature’s considered views about the notwithstanding clause. Instead, certainly the former and arguably the latter emanate from the executive rather than the legislature.

Without meaning to, SSM give away the notwithstanding clause defenders’ sleight of hand: while they denounce those who have but “a limited regard for the legislature’s capacity to reason about rights”, they are, in reality, apologists for executive power. Unsurprisingly, they repeatedly speak of the government, not the legislature, invoking the notwithstanding clause. Earlier, they cheerfully note that Premier Doug “Ford’s government controlled the legislature, and so the bill” that expanded the censorship of political advertising before elections “passed with little difficulty”. This all is, of course, of a piece of the “common good” movement’s embrace of executive and administrative power elsewhere. Professor Vermeule, for instance, is an advocate of “law’s abnegation”, as the title of one of his books has it, in the face of the administrative state. SSM themselves defend approaches to legal interpretation that would empower administrative decision-makers instead of holding them to the limits enacted by legislatures.

This power, moreover, is an unbridled one. Recall that, contrary to SSM’s insistence on (legislative) reasoning about rights, the Ontario government advanced no reason at all to justify its expansion of political censorship. To repeat, the Superior Court did not disagree with the government’s justification or rule that it was insufficiently supported by evidence ― though it’s worth pointing out that there never has been any evidence that the integrity and fairness of Canadian elections were compromised by the lack of a year-long gag on the civil society, or even by the absence of the much more modest restrictions upheld in Harper. The Harper majority specifically held that evidence was unnecessary ― a reason, among others, why Harper is one of the Supreme Court’s worst decisions of all time.

Be that as it may, the Working Families court found that there was no justification at all for limiting the freedom of expression of civil society groups for as long as the legislature had. For all that SSM claim to regard “law as a work of reason”, for all their insistence that “[t]hrough a prudent exercise of reason, the law-maker is free and apt to make a practical judgment in choosing among the many alternatives, the many legitimate and reasonable possibilities”, the law they actually extol is an unreasoned power-grab by the executive. By asking us to accept it in the name of reason, SSM show that this rhetoric is just a spice intended to mask the insipid taste of their actual position.

And, for all their contempt for legal positivism and posturing as the heirs to the natural law tradition, SSM are, in truth, asking us to accept the authority of law simply because it has been enacted by the state. They deprecate as simple-mindedly positivistic the view of “legal rights as solely the emanation of judicial decisions”, so that “a Charter right is effectively nullified if the legislature derogates from judicial review via the notwithstanding mechanism”. (SSM never say, of course, who actually holds these views.) For them rights, being emanations of the natural law, exist even if they cannot be enforced through the courts.

But individuals must accept the legislature’s ― or rather, as we have seen, the executive’s ― specification of these rights, even when, as in the case of Ontario’s censorship regime and its use of the notwithstanding clause, the legislature manifestly failed to turn its mind to the right in question. No other reason than the legislature’s authority, and the common good constitutionalists’ naïve believe in its ability to reason, is necessary. And of course, like all notwithstanding clause apologists, SSM trot out the historical fact that it is “part of the Charter and the political settlement that made possible the constitutional entrenchment itself”, as if that can legitimate political actors resorting to it. But that is only so on a nakedly positivist view, where the legality of something is sufficient warrant for its legitimacy.

As co-blogger Mark Mancini and I have previously suggested here and here, SSM’s embrace of common good constitutionalism is superfluous at best, and actively pernicious at worst. If is superfluous if it only serves to provide a baroque vocabulary for warmed-up arguments for in favour of political power and against judicially-enforceable individual rights. It is pernicious if they really mean to embrace the most reactionary views associated with, and sometimes openly embraced by, their ideological fellow travellers.

On the whole, their Ius et Iustitium post is evidence for the former possibility. Little if anything in it could not have been said, and has not been said, without the “common good” sauce. But even stripped of this rhetoric, the argument remains distasteful enough. Citizens ought to defer to the choices executive branch officials, so long as they have been laundered through supine legislatures, because these legislatures in theory could have ― and it doesn’t matter that they actually haven’t ― engaged in reasoned deliberations about rights. Calling something an exercise of reason directed at the common good does not make it so. Tripe is tripe, and a power grab is a power grab.

Telling People Whom to Vote for

An illiberal community seeks to dictate its members’ votes. How can, and should, the law respond ― and quite how different are liberal democracies anyway?

When it comes to election campaigns, where does permissible ― and perhaps even laudable ― persuasion end, and deplorable ― perhaps even illegal ― manipulation or indeed coercion begin? This is a fraught question, as a recent story by Sally Murphy for Radio New Zealand illustrates.

The story concerns what seems to be an totalitarian and abusive fundamentalist religious community, whose leaders seek to dictate not only how members will live, but also how they will vote:

Former members of Gloriavale Christian Community say people still there do not have the freedom to vote for who they want in the general election. … [T]hose inside don’t have free access to the internet or news sources and are told as a collective who to vote for. … One former member … told RNZ Gloriavale leaders would choose which politicians would come and talk to the community before an election. 

“They would talk about their policies and what they would do for us then when they left there would be a discussion, but it was usually only a couple of the leaders who would talk,” she said. “They would say we like this party because of this policy and that we should all vote for them because it’s best if we vote as a collective.” 

It seems fair to infer that current members are likely not to feel free to cast a vote at odds with the preferences of their leaders. But does that mean that something untoward or illegal is going on, and further, that something can, or should, be done?

Consider, first, existing election law. (I am leaving out the charities law aspect of this issue, mostly because it’s not my area of expertise. For a discussion of the restrictions on charties’ ability to engage in politics in the Canadian context, see this guest post by Benjamin Oliphant; and for a broader discussion of the tensions at work in the regulation of charities, this guest post by Kathryn Chan.) Section 218 of the Electoral Act 1993 makes it an offence and a “corrupt practice” to

make[] use of or threaten[] to make use of any force, violence, or restraint, or inflict[] or threaten[] to inflict … any temporal or spiritual injury, damage, harm, or loss upon or against any person, in order to induce or compel that person to vote for or against a particular candidate or party … or on account of that person having voted for or against a particular candidate

But ― going by the statements quoted in Ms. Murphy’s story ― no threats are being made in relation to voting specifically. Gloriavale members are told to vote a certain way, but not actually threatened with reprisals if they do not. Besides, as the Electoral Commission points out, the secrecy of the ballot ought to mean that all voters, including Gloriavale members, can “express their preferences free of outside influence or coercion”.

Beyond threats, the regulation of the persuasion of voters focuses (in various ways) mainly on spending and to some extent on the use of mass media, especially broadcast media. Private, face-to-face exhortation is not targeted, and it would be absurd if it were. Would we want political conversations within families or among friends and co-workers to be subject to regulation? I should think not, even though some of these conversations may be emotionally charged, and people may be uncomfortable, or even distressed, at the idea of going against the wishes or preferences of those close to them. Again, the primary remedy for such situations is ballot secrecy, with section 218 outlawing outright threats.

If the Gloriavale leadership is not breaking election law, is it nevertheless acting immorally in seeking to influencing the members’ votes, and should the law be changed? Again, in relation to voting specifically, it’s not obvious to me that a wrong is being done. All sorts of people and entities tell us to vote one way or another. I don’t think that they necessarily wrong us just by doing so ― even if these people are close to us and may be reluctant to offend or contradict them. Just as it would be absurd to have legal rules regulating political discussions among friends, family members, or other close associates, I think a moral rule to this effect would be contrary to widely held views of both of a good life and of good democratic citizenship. The latter, in particular, surely permits people to urge others to vote in ways they consider to be better for the community.

What makes the Gloriavale situation disturbing is the broader atmosphere ― the habit of obedience and the limitation of alternative sources of information and opinion imposed on its members. People who tell us, even quite forcefully, that we should vote one way or another do not wrong us if the choice is ultimately ours. People who keep us from making an informed choice wrong us even if they do not impose their own preferences. Imagine, hypothetically, that the Gloriavale leadership did not tell the members how to vote. To the extent that they are simply denied information from the outside world, the members would have no idea, and would not be in a position to make a more meaningful choice than they are now, and those responsible for putting them in this position are to blame.

However, election law is not the remedy for such cases. The challenge of illiberal and authoritarian communities within their midst is not an easy one for liberal societies, but to the extent it can be solved at all, the solution has to be at a rather more fundamental level. Perhaps ironically, though, extreme examples like Gloriavale can help us reflect on the fact that liberal societies themselves are not entirely innocent of trying to restrict the information and choices available to their members.

Hard restrictions are, admittedly, rare. Yet not non-existent. In New Zealand, the Classification Office, headed by a Chief Censor, is empowered to ban publications in various media. While that outfit’s website’s proclaims that its enabling legislation “does not regulate political speech, the expression of opinions, or ‘hate speech'”, it has notoriously banned the Christchurch shooter’s “manifesto”, which is obviously an example of ― horrible ― political speech. As Ilya Somin has shown, reading it, in all its gruesomeness, is actually instructive. But New Zealanders are not permitted to do it, because the Chief Censor, on his own motion, decided that he knew better. In some other democratic countries, especially in Europe, political choices can be restricted by the authorities banning political parties deemed opposed to democracy or the existing constitutional order.

Of course, these are extreme examples. There is no equivalence between excluding some outlier political options while preserving a wide range of choice and excluding all options but one. Arguably that the most important thing about democracy is not the ability to vote for one’s preferred agenda ― which constitutional constraints or the vagaries of the electoral system, not to mention a shortage of people who agree with it and are willing to run for office ― might make impossible, but simply the ability to make some kind of choice, and so to throw the bums out from time to time. Still, the censorship impulse has a common foundation in both cases ― the distrust of people’s ability to make acceptable choices, and a confidence in one’s ability to choose on others’ behalf.

And softer, more insidious ways of shaping the range of choices available to voters are common. How do teachers are university professors speak about political views outside the mainstream ― or outside what they perceive as the mainstream? How, if at all, do the media cover unorthodox politicians, at least those who do not also happen to be celebrities? Are the above-mentioned regulations of spending on election campaigns structured so as to favour established parties ― as they are in New Zealand, for example, with the allowed spending on broadcast advertising dependent, in part, on a party’s share of the vote in the previous election? To ask these questions is not, by itself, to advocate for root-and-branch reform of the education system, the demise of the legacy media, and complete deregulation of electoral campaigns. But here again the effects of seemingly disparate and often well-intentioned policies and practices commonly followed in liberal democracies are a little less different from those of the practices of demonstrably illiberal communities than we might be quite comfortable with.

The difficulty of ensuring that all voters, including those who happen to belong to heterodox and illiberal communities, are able to take a meaningful part in an election if they wish to should not stop us from trying. Features of the electoral process that help facilitate meaningful participation and might strike us as obvious today, such as the secret ballot, did not always exist: they had to be invented, and the law had to be changed to implement them. One should of course be wary of unintended consequences, including those of well-meaning but excessive regulation. But perhaps there are ways to make things easier for members of Gloriavale and others caught in similar situations, without introducing unnecessarily intrusive laws. But as we look for such solutions, we should remember that existing laws and practices constrain the range of political choices available to all citizens, and that some of them have effects that differ in degree, but perhaps not in kind, from those of the impositions at Gloriavale and elsewhere. Not all authoritarians in our midst are content to run cults.

The System Is Working

Environmentalist groups have a point when they say they are being muzzled by Elections Canada; trouble is, that’s exactly how the law is meant to work

As the media reported earlier this week, environmentalist groups are angry at Elections Canada, which has warned them that spending money to raise awareness of climate change in the run-up to the coming federal election would subject them to the rules on “third party” participation in election campaigns. Many are feeling that they will be required to keep quiet during the campaign, which rather defeats the purpose of being advocacy groups. Even the BBC has a story on this.

For its part, Elections Canada has issued a response claiming that the Canada Elections Act doesn’t prevent advocacy groups from advocating, so long as they register if they spend $500 or more and comply with the spending cap. Elections Canada adds that the registration requirement “leads to increased transparency” and has been in place “for nearly 20 years”. Helpfully, I suppose, the statement concludes with an acknowledgement that the rules “can be complex”, and Elections Canada is happy to answer questions about them.

The rules are indeed somewhat complicated, as I explain below. But the bottom line is simple enough. Despite the officials’ protestations, NGOs ― be they environmentalist or other ― have a point when they say that they are being muzzled. To some extent, that’s what the Canada Elections Act is designed to do; to an even greater extent this might be an unintended consequence of the Act’s pursuit of transparency, but an entirely predictable one. The issues are well known; I, for one, raised them in my statement to the House of Commons Select Committee that considered the latest round of amendments to the Canada Elections Act. The only surprising thing is the degree to which people still end up being surprised when problems of sort arise.

The Canada Elections Act‘s regulation of political spending is predicated on the idea that attention during election campaigns should be focused on politicians ― individual candidates and political parties, especially parties. Parties, if they run candidates in all ridings, are able to spend tens of millions of dollars on advertising ― which they are entitled to buy at favourable rates, in addition to an allowance of free airtime. Non-politicians ― that is, individuals, labour and student unions, corporations, and NGOs ― are known as “third parties” in the election law jargon and, as I explained here, their participation in electoral debates is viewed as anomalous, indeed suspicious, and is strictly limited.

One set of limits concerns the amounts of money third parties are allowed to spend, which are only a small fraction of the spending allowed political parties. The Supreme Court has upheld the limitation of third party spending during election campaigns, notably in Harper v Canada (Attorney General), 2004 SCC 33, [2004] 1 SCR 827, although there is good reason to be critical of that decision, which I have even rated as one of the worst in the last fifty years. (As I noted here, the High Court of Australia was also quite skeptical of Harper in a recent decision.) Last year, Parliament enacted further limits that apply even before the formal campaign begins, and their constitutionality has not yet been tested; Harper, in my view, does not dispose of the question.

In addition to spending limits, “third parties” are also subject to onerous registration and reporting requirements. Some of these are the cause of the latest dust-up. Specifically, Division 1 of Part 17 of the Canada Elections Act imposes such requirements on “third parties” that incur more than $500 of expenses on, notably “partisan activities” and “partisan advertising” during the “pre-election period”, which begins on June 30 of the year for which a fixed-date election is scheduled and ends with the start of the election campaign. During the election campaign itself, governed by Division 2 of Part 17, “election advertising”, as well as “partisan activities” count for the spending thresholds that can trigger registration and reporting requirements.

The definitions of “partisan” and “election advertising”, found in section 2(1) of the Canada Elections Act, are very broad. The former term “means the transmission to the public by any means during a pre-election period of an advertising message that promotes or opposes” a party or a candidate, further defined in section 2(7) as “naming”, “identifying” (“including by … logo” or picture, as the case may be, and “providing a link to an Internet page that” names or identifies the party or candidate. “Election advertising” includes the same things as “partisan advertising”, but also “taking a position on an issue with which a … party or candidate is associated”, even without naming that party or candidate. Since issues with which no candidate or party “is associated”, come election time, are about as common as colour pictures of a Maple Leafs Stanley Cup parade, the definition of “election advertising” encompasses pretty much any advertising that has anything to say on matters of government or policy.

Now, some means of communicating with the public are exempted from these definitions. In particular, the exemptions cover anything that the media will print or broadcast without charge to the speaker ― things like quotes in news items, interviews, and op-eds. Also exempt are organizations’ communications with their members, shareholders, or employees, as well as “the transmission by an individual, on a non-commercial basis on the Internet, of his or her personal political views”. Note, though, that on its face the latter exemption doesn’t cover ― indeed, it rather pointedly excludes ― a group’s or an organization’s online communications, even if not paid for (for example, tweeting under the organization’s handle). And of course, any communication that the media are not interested in carrying free of charge will count as an advertising. In effect, for groups and organizations, the media are the gatekeepers of their ability to communicate with the public without having to register as “third parties”.

So what’s the big deal about registration? Well, although you won’t know it from the Elections Canada statement linked to above, registration doesn’t just mean filling out a form. There are a number of other requirements. To begin with, unions and corporations cannot register before their board has adopted a resolution authorizing them to incur expenses on “partisan” or “election advertising” (sections 349.6(5) and 353(5) of the Canada Elections Act). All “third parties” are also required to have a “financial agent” who will be responsible for collecting money to be spend on “partisan” or “election advertising” and for spending it (sections 349.7 and 354). These transactions must be done through a separate bank account (section 358.1) After the election is over, a detailed report on the money collected, advertising taken out, and costs incurred must be filed (section 359). And this is not all. Those “third parties” that spend more than 10,000$ are also required to file interim reports during the course of the election campaign and, most significantly, to appoint auditors (section 355) and file the auditor’s report on their spending (section 360).

Needless to say, this is all quite costly, at least in time, but also ― especially for those third parties that spend more 10,000$ ― in money. Big trade unions, whose budgets are extracted from workers who don’t get a say on whether to contribute or on how the money is spend, may not be especially troubled by these costs. But for NGOs, whose income comes from voluntary (albeit taxpayer-subsidized) donations, and which need to be much more careful about how they spend it, compliance with the Canada Elections Act may be too expensive. From their perspective, the sensible if unfortunate thing to do may well be to keep quiet for the duration of the election campaign, or even starting with the beginning of the pre-campaign period.

This means that for a period of almost four months preceding the election ― the period when the most people pay attention, even if it’s still sporadic and fragmentary attention, to political and policy issues ― civil society organizations may indeed be prevented from expressing their opinion about politicians, except to the extent that the media will let them. Again, the bigger and better-known you are, the less of a problem this may be for you. Smaller groups, whose views are (naturally and fairly) of less interest to the media, will find it more difficult to get across to the voters. The more unusual voices, in other words, are the ones who are the most at risk of being silenced ― in effect if not, perhaps, in intent ― by the Canada Elections Act.

And of course even for larger groups, having to pass through the media means sound-bite-sized interventions have a much better chance of getting across to the voters than anything more serious. Say that a politician or party is anti-environment, or pro-worker, or something equally inane, and the media may well pick it up. But they’re not going to run a detailed report card assessing the competing parties’ platforms on some issue ― but publishing it on an NGO’s website, let alone running it as an advertisement would mean having to comply with burdensome registration and reporting requirements under the Canada Elections Act.

No wonder, then, that environmentalists are feeling muzzled and frustrated. And of course groups pursuing other agendas may be feeling that way too ― or may come to feel that way when the occasion arises. They have more than a little justification. And they shouldn’t be the only ones feeling wronged. The voters should be too. You may not miss the presence of a particular set of activists in the election campaign, but the rules that silence them silence the activists on your side too. You may not be all that interested activists generally have to say, but you should be interested in politicians’ feet being held to the fire.

The ostensible rationale for registration and reporting requirements is that they serve to promote transparency, in addition to assisting in the enforcement of spending limits applicable to “third parties”. It is on that basis that the Supreme Court upheld those requirements that apply in the course of the election campaign ― although not those applicable in the pre-campaign period, which weren’t yet in the Canada Elections Act ― in Harper. Yet one needs to weigh the value of transparency against the costs that its pursuit imposes on those subject to the Canada Elections Act ― and, as I have just explained, on the voters who are being deprived of important contributions to the electoral debate.

The Harper majority’s analysis on this point was quite perfunctory. There is no real discussion of compliance costs and their deterrent effects. Instead, the majority is content to baldly assert that “[t]he appointment of a financial agent or auditor is not overly onerous. Rather, it arguably facilitates the reporting requirements.” [145] Even worse, the majority did not at all consider what I think is the crucial issue: the thresholds at which the registration and reporting requirements kick in. All it said was that the requirements “vary depending on the amount spent on election advertising”. [145] Yet one can accept the principle of imposing such requirements on heavy spenders while also acknowledging that the existing rules go much too far.

In New Zealand, “third parties” are not required to register until they spend NZ$13,200 (ca. C$11,000); more detailed reporting requirements only apply once a “third party” spends NZ$100,000. (Even then, third parties aren’t peremptorily required to provide an auditor’s report, although they may be asked to do so.) These strike me as rather more reasonable figures than those in the Canada Elections Act, though even they should probably be multiplied several-fold to account for the fact that New Zealand’s population is only a small fraction of Canada’s.

It is difficult to believe that a “third party” spending a few thousand, or even tens of thousand of dollars is going to have any substantial impact on an election by itself. At most, it may be successful enough in getting other people ― voters, media, or politicians ― to discuss the issues it is raising. It is this discussion, rather than anything published on an NGO’s website or even a Facebook ad, that might, conceivably albeit theoretically, matter. In the abstract, this discussion might be enriched by more disclosure. In practice, the very real costs of the disclosure requirements end up preventing the conversations from happening at all. I fail to see how the voters benefit from this.

As Elections Canada points out in its response to the environmentalist groups, the “advertising during the election period has been subject to the Canada Elections Act for nearly 20 years”. This is true. (As noted above, rules on advertising in the pre-election period are new.) For about half of this time, it has been known, at least to those who study these things, that the rules tend to hobble not business interests, but labour unions and civil society groups. Colin Feasby wrote about this in 2010; I did (in the context of Québec elections, which are subject to similar but even more draconian rules) in 2012; also in 2012 Tom Flanagan came out in support of rules like those in the Canada Elections Act, whose enactment he had opposed, with the declared intention to muzzle unions; I updated Dr. Feasby’s findings in an article published in 2015. And in my statement to the House of Commons Standing Committee on Procedure and House Affairs when it was studying amendments to the Canada Elections Act last year (which, among other things, introduced restrictions on “third parties” in the pre-campaign period) I specifically mentioned both the registration and reporting requirements’ tendency to muzzle civil society, and the needless low threshold at which these requirements apply. Needless to say, that had no effect on the resulting legislation.

Yet at every election the impact of restrictions on “third parties” seems to surprise. It happened in Québec in 2014, when the Chief Electoral Officer tried censoring a short documentary a group of citizens had produced to oppose the election of the Parti québécois and the enactment of its “values charter”. Eventually, the Chief Electoral Officer changed his mind; but he was wrong to do so. It happened again in Québec in 2018, now with environmentalist groups being targeted. And now it’s happening at the federal level. The system, one might say, is working. It was designed to shut down political debate not dominated by politicians or the media. That’s what it’s doing.

It will be obvious that I don’t think it’s a good system. Like the National Post’s Chris Selley, I think the rules need to be changed. Whether any restrictions on political spending are justified is debatable but, as noted above, one can accept the premises of Canada Elections Act and still support relaxing its requirements a great deal. Ideally, the next Parliament will take up the issue. But there is also room for litigation. Certainly rules on pre-campaign spending, whose constitutionality has not yet been tested all the way to the Supreme Court can be challenged. But perhaps even the registration and reporting rules upheld in Harper could be attacked, provided that the courts are forced to consider solid evidence of their pernicious effects.

Australia 1:0 Canada

Canadians have much to learn from the Australian High Court’s take on election spending limits for “third parties”

The High Court of Australia has just delivered Unions NSW v New South Wales [2019] HCA 1, a decision that should be of interest to readers who are concerned with freedom of expression in the electoral context ― a topical issue in Canada, given the recent imposition of further restrictions in this area by the recently enacted Bill C-76. The decision resulted from a challenge by a number of labour unions to New South Wales legislation that reduced the maximum amount a “third party” ― that is anyone not a candidate at an election or a political party ― is allowed to spend on campaigning in a nearly-six-month period preceding an election, from 1,050,000AUD (jut under a million Canadian) to 500,000AUD. The High Court unanimously held that the legislation was contrary to the implied freedom of political communication, which it had previously read into the Australian constitution‘s provisions requiring “representative” government.

The plurality judgment, by Chief Justice Kiefel and Justices Bell and Keane, finds that the third party spending limits are unconstitutional. That they restrict the ability to communicate is not in dispute. And while the plurality is prepared to assume that these limits are imposed for the legitimate purposes of levelling the campaigning playing field and preventing the wealthy from “drowning out” the voices of the less fortunate, they are not justified. Experts consulted prior to the enactment of the legislation provided no particular justification for recommending that the then-existing spending limits be reduced. A Parliamentary committee, however, recommend that the legislature look into the actual spending needs of third parties, and this was not done either. As a result, there is no reason for saying that the reduced limits are “reasonably necessary”.

Justice Gageler agrees with the plurality’s disposition of the case. He is persuaded of the legitimacy of the state’s pursuit “of substantive fairness in a manner compatible with maintenance of the constitutionally prescribed system of representative and responsible government”. [91] This might, in principle, justify much lower spending limits for third parties, which campaign on single issues, than for parties that must address a broad range of issues in their quest to form a government. However, “[i]t is not self-evident, and it has not been shown, that the cap set in the amount of $500,000 leaves a third-party campaigner with a reasonable opportunity to present its case”. [101] Absent such a showing, the restriction on the freedom of communication is not justified.

Justice Nettle’s conclusion is similar. He accepts the legitimacy of the objective of creating a level electoral playing field ― one on which political parties will be primary players ― and agrees that a legislatures may from time to time review the measures it takes to ensure fairness, including by lowering spending caps previously enacted. However, there must be a justification for whatever measures it takes from time to time. Such a justification is missing in this case. Although it was recommended that more evidence on the needs of third parties be collected, “for reasons which do not appear, that recommendation went unheeded. It is as if Parliament simply went ahead … without pausing to consider whether a cut of as much as 50 per cent was required”. [117]

Justice Gordon, like the plurality, assumes that restrictions on third party spending pursue a legitimate purpose, which she characterizes as the privileging of political parties and candidates. However, in the absence of evidence about the actual need for restrictions set at their current level, “the Court … cannot be satisfied that the level of the expenditure cap is reasonably appropriate and adapted to achieve the asserted constitutionally permissible end”. [150] It was for the State to show that the restriction it seeks to impose was justified, and it has not done so.

For his part, Justice Edelman considers that the reduction in the spending limits imposed on third parties, even as the limits imposed on political parties rose, cannot be explained by the purposes of maintaining a fair and corruption-free electoral system. Rather, it must have had an “additional purpose”, which “was to ensure that the voice of third-party campaigners was quieter than that of political parties and candidates”. [159] In other words, the reduction’s aim was “to burden the freedom of political communication of third-party campaigners”. [160] Justice Edelman considers that, although laws that rely on the relative silencing of some views in order to ensure that all can be heard are legitimate, to aim only at silencing some voices “is incompatible with the maintenance of the constitutionally prescribed system of representative and responsible government”, and legislation so motivated is “invalid”. [160]

Needless to say, I am not qualified to comment on whether the High Court is correct as a matter of Australian law. What I can do is compare its decision with that of the Supreme Court of Canada in Harper v Canada (Attorney-General), 2004 SCC 33, [2004] 1 SCR 827, which addressed much the same issues. (Readers will recall that I am not a fan of Harper, to put it mildly, and included it in my list of the Supreme Court’s five worst decisions of the last half-century in this blog’s recent Twelve Days of Christmas symposium.)

This is most obviously so on the issue of deference to the legislature on the issue of the appropriateness of a limitation of the freedom of (political) expression, and the evidence required for the government to make this case. The Harper majority insisted that courts should approach legislative choices with deference. In its view, “[t]he legislature is not required to provide scientific proof based on concrete evidence of the problem it seeks to address in every case”, and that “a reasoned apprehension of … harm” [Harper, 77] is sufficient to restrict fundamental freedoms protected by the Canadian constitution.

This approach is explicitly rejected in Unions NSW. While the Australian judges avoid directly criticizing the Harper majority, both the plurality opinion and Justice Nettle explicitly side with the “strong dissent” of Chief Justice McLachlin and Justice Major (joined by Justice Binnie). The plurality takes a dim view of the submission “that Parliament does not need to provide evidence for the legislation it enacts [and] is entitled to make the choice as to what level of restriction is necessary to meet future problems”. [44] When legislative choice are made in a way that burden the freedom of political communication, they must be justified. Similarly, Justice Gageler speaks of the need for a “compelling justification”, and insists that “[i]f a court cannot be satisfied of a fact the existence of which is necessary in law to provide a constitutional basis for impugned legislation, … the court has no option but to pronounce the legislation invalid.” [95] Justice Gordon insists that “the Court must … be astute not to accept at face value the assertion that freedom of communication will, unless curtailed by a reduction in the cap to $500,000, bring about corruption and distortion of the political process”. [148]

Another point of contrast between Harper and Unions NSW is the treatment of the so-called “egalitarian model of elections” designed in part to favour the interests of political parties and candidates over those of the civil society groups, disparagingly consigned to the status of “third parties”. According to Harper, election campaigns must focus attention on parties and candidates, including by ensuring that any other participants in the public debate, except the media, will behave unobtrusively. By contrast, the plurality opinion in Unions NSW explicitly rejects the submission that candidates and parties deserve preferential treatment, advanced in part on the basis that elections are “not a choice between ideas, policies, views or beliefs except insofar as such choice may be reflected in the electoral choice between candidates”. [39] Rather, the plurality says, “ss 7 and 24 of the Constitution guarantee the political sovereignty of the people of the Commonwealth by ensuring that their choice of elected representatives is a real choice, that is, a choice that is free and well-informed” [40] ― including by third parties. Justice Gageler, of course, takes the contrary view on this point. Justice Edelman’s position is more complex. He explicitly endorses “a Rawlsian, egalitarian model” [178] in which spending limits prevent some speakers from “drowning out” others. However, he also considers that it is not legitimate to target particular speakers for silencing apart from such an anti-drowning out purpose.

A last difference between Harper and Unions NSW worth highlighting is recognition by Justice Gageler of “the propensity of an elected majority to undervalue, and, at worst, to seek to protect itself against adverse electoral consequences resulting from, political communication by a dissenting minority”. [66] Justice Gageler refers to prior cases where the risk of a government legislating to limit political competition the better to maintain itself in office was explicitly adverted to. Such legislation, he notes, is incompatible with presuppositions of the Australian constitutional order. Although he finds that, in this case, “[t]here is no suggestion of abuse of incumbency” [85] by one party against others, this clear-eyed position is in contrast to that of the Harper majorityr, which ignored the possibility that incumbent governments favour legislation that excludes “third parties” from electoral campaigns in order to avoid unpleasant criticism and so reduce the odds of losing power.

There are more interesting things in the Unions NSW decision than I have room to discuss in this post. For example, Justice Gageler’s comments about the role courts in finding facts that are relevant to deciding whether a statute is constitutional are in contrast to the position of the Supreme Court of Canada in cases such as Canada (Attorney General) v Bedford, 2013 SCC 72, [2013] 3 SCR 1101, and should be very nutritious food for thought for those who are skeptical of the Bedford requirement of deference to trial judges. Justice Eledman’s comments on identifying statutory purpose (and in particular the role of general statements of purpose in the legislation) are also very interesting.

Overall, based on this one decision, I think that Canadians have a great deal to learn from Australians. Admittedly, the length of the High Court’s decisions is a deterrent ― Unions NSW is about 85 pages long, and I take it that it’s pretty short by Australian standards. That’s the cost of so many judges delivering full individual reasons. But the upside is that interesting ideas don’t get swept under the carpet in the process of getting to a set of reasons many judges can sign onto. I’m not saying the Supreme Court of Canada should go back to having each judge deliver his or her own reasons (though I wonder sometimes) but, at any rate, reading the Australian decisions may well be worth our while. In particular, the willingness of the Australian judges to keep a legislature accountable for imposing limits on the freedom of political expression without justification is a welcome reminder that their Canadian counterparts can do much more to protect individual rights in the electoral realm, and elsewhere.

Civics, Feelings, and Politics

Expatriates’ alleged lack of connection to particular ridings is not a good reason to disenfranchise them

When it held, in Frank v Canada (Attorney-General), 2019 SCC 1 (summarized here), that denying the franchise to Canadians who have been resident abroad for more than five years is unconstitutional, did the Supreme Court go with “feelings over civics”? Did it decide the case in a way that ignores the fact that Canadians vote not for national parties but for candidates in local constituencies, to which expatriates are not meaningfully connected, even if they maintain, as the Court’s majority said, a “profound attachment” to Canada? Over at Routine Proceedings, Dale Smith argues that that’s precisely what the Court did. I disagree.

As Mr. Smith sees it, “five of seven justices of the Supreme Court failed to properly understand the importance of constituency-based democracy”. He also faults the government’s lawyers “for not making the case adequately either”, “and virtually all of the commentary” on Frank, including presumably my comment, for ignoring the issue. Yet in his view, it ought to have been a decisive consideration:

[W]e vote for local representatives. We don’t vote for parties, or party leaders, no matter what we may have in mind when we go into the ballot box – we mark the X for the local candidate, end of story. For an expat, it’s not the connection to Canada that should be at issue – it’s the connection to the riding, because that’s how we allocate our votes.

One might, of course, reproach the government lawyers for failing to emphasize this particular rationale for disenfranchising Canadians abroad. The Frank majority, even on this view, is blameless, because it wasn’t at liberty to sustain the disenfranchisement on the basis of a justification that the government did not even put forward. Section 1 of the Canadian Charter of Rights and Freedoms provides that limits on Charter rights, including the right to vote, must be “demonstrably justified” ― and as the Supreme Court has long held, it is the government that must carry out the demonstration. But there are other reasons, based in both what we might (loosely) call civics and feelings, as well as some realism about politics, that mean that, had the government chosen to make connection to the riding as the hill its case would die on, this case would be every bit as dead as it now is.

Start with the civics. Mr. Smith is quite right that, in point of law, we vote for local representatives, not for national parties or their leaders. Whether this ought to matter as much as he suggests, I will discuss below, when I turn to politics. But it’s important to consider a couple of other legal issues.

First, though there seems to be a good deal of confusion or even obfuscation on this point, the Canada Elections Act already takes care of the need for a connection between a Canadian voting from abroad (who may be a short-term expatriate, a long-term one previously allowed to vote, such as a diplomat’s family member, or a newly-enfranchised long-term expatriate). Paragraph 223(1)(e) provides that, when applying to be registered as an elector resident outside Canada and requesting to vote by special ballot, a would-be voter must provide the Canadian address to which his or her vote will be tied. Once the choice has been made, section 224 prevents the voter from changing it. This prevents forum shopping, as it were, and seems a sensible regulation.

Now, there is a range of options for the prospective voter from abroad to choose from:

the address of the elector’s last place of ordinary residence in Canada before he or she left Canada or the address of the place of ordinary residence in Canada of the spouse, the common-law partner or a relative of the elector, a relative of the elector’s spouse or common-law partner, a person in relation to whom the elector is a dependant or a person with whom the elector would live but for his or her residing temporarily outside Canada.

It has been put to me that the breadth of this range is excessive and gives the elector too much choice. If Parliament agrees, it can eliminate some superfluous options by legislation; this should not be constitutionally problematic. But I don’t think that Parliament should do this. On the contrary, giving the voter the ability to tie his or her vote to a former residence or a family member’s one makes it more likely that the elector will choose to vote at the particular place in Canada to which he or she is feels the strongest connection, which will not be the same for all expatriates, and which each voter is much better positioned to figure out when registering than Parliament when legislating.

Second, one must keep in mind that when it comes to voters in Canada, the law does not require any sort of evidence of a connection between the voter and his or her riding other than the fact that the voter resides there. Perhaps that’s because residence is simply deemed to be determinative of the community to which the voter belongs. But this seems a very rough assumption, especially in today’s urbanized world, in which many ridings are quite compact and the boundaries between them, fluid. A voter might be live in a bedroom community or a residential neighbourhood, but work in a downtown in a different riding, and perhaps have other attachments in yet a third one. It is, to say the least, not obvious which of these the voter is genuinely connected to. Residence, arguably, is only the most easily administrable way of sorting voters into ridings (both at the point of counting them through the census and at the point of registering them), simply because it tends to be more stable than other connections. As Chief Justice Wagner, writing for the Frank majority, put it, “residence can best be understood as an organizing mechanism for purposes of the right to vote”. [28] It is nothing more than that.

This brings me to what Mr. Smith might calls “feelings”. He and others who defend the disenfranchisement of Canadians abroad are very quick to demand that expatriates meet conditions that are not imposed on other Canadians to qualify for the franchise. Whether it be some subjective connection to a riding or to Canada as a whole, or knowledge about the local state of affairs, or tax liability, or subjection some undefined but substantial number laws, not all residents will meet these conditions that are said to justify denying the franchise to expatriates. But no one thinks to inquire into whether they really do, and no one, I’m pretty sure, would accept (re-)introducing tests of this nature into our election laws. Expatriates are the only people whom people judge on such criteria.

Indeed, it is not so much a judgment as prejudice. Expatriates are simply assumed to fail such tests ― and arguments to the contrary are dismissed as “feelings”. Mr. Smith guesses that Canadians who live abroad cut themselves off from communities where they used to live, or have family, or intend to return (or all of these things). Why? My personal experience, for what that’s worth, is that I keep up with the news from Québec and Montreal (and occasionally write on Québec-specific issues), more than from other provinces. Do I specifically track the news for Notre-Dame-de-Grâce-Westmount, where my parents live and I will vote in October, if I can be bothered? Not particularly, but then again, I wouldn’t even if I actually lived there. To say that I’m not a suitable voter for this reason would be applying a groundless double standard.

And speaking of double standards (and, I suppose, of civics), it’s worth noting that pursuant to section 222 of the Canada Elections Act some long-term expatriates are already allowed to vote: namely, members and employees of the Canadian forces, federal provincial public servants, employees of “international organization[s] of which Canada is a member and to which Canada contributes, as well as anyone who “lives with” such voters. The rationale for this is, presumably, that all such persons ― not just public servants, mind you, but their family members too ― are deemed to maintain a connection with Canada that other expatriates lack. Yet even assuming that this is so, is it remotely plausible that such persons (who, if anything, probably tend to be more mobile than the average voter even when they live in Canada) maintain their special connections to their home ridings? I really don’t think this is plausible, and so, the invocation of the riding connection as a justification for disenfranchising some, but not all, expatriates is another sort of unwarranted double standard.

Let me finally turn to politics ― and, specifically, to the need to be realistic about it. If we want to understand the rules of elections and government formation in Canada, we must keep in mind that each voter only casts a ballot for a local representative, not (directly) for a party or Prime Minister. But if we want to figure out whether Parliament is justified in preventing a person or a class of persons from voting, I don’t think it makes sense to pretend, as Mr. Smith asks us to, that this is all that matters. The reality, as he more or less acknowledges, is that what we “have in mind when we go into the ballot box” ― or at least the voting booth, for the less acrobatic among us ― very much has to do with parties and, especially, their leaders, for most voters.

Political parties themselves know this. The big ones tried to prevent to keep the small ones from getting their names on ballot papers, until the Supreme Court wisely put an end to that in Figueroa v Canada (Attorney General), 2003 SCC 37, [2003] 1 SCR 912, because candidates not identified with parties get fewer votes. They give pride of place to party names, logos, slogans, and leaders in their advertisements. They make sure their MPs have lookalike websites in party colours. Local candidates are often little more than props for a leader’s tour. I’m too lazy to look for the relevant research (if it exists) right now, but as a not-so-wild guess, I’m inclined to think that many voters don’t even remember the name of their local candidate when they go vote. This may be regrettable, but the parties themselves have ensured that it doesn’t matter; what does matter is the party identification on the ballot paper.

One key reason for this is that election campaigns are largely national events, not local ones. (By way of thought experiment, imagine we didn’t hold simultaneous general elections, but renewed the House of Commons with staggered elections, one riding at a time. Our politics may well be quite different ― and more local. But of course we don’t do that.) The centrality of leaders’ personalities to election campaigns makes this unavoidable, and an even starker phenomenon than in the past. But even to the (limited) extent that voters are preoccupied with actual issues rather than personalities, the issues are largely national in scope. This is perhaps especially the case in federal elections, since Parliament’s powers are, by design, largely those that concern the country as a whole. Admittedly Parliament doesn’t always keep to its jurisdiction. Even when it does, Justice Rowe points out in his concurring reasons that “federal policy can impact different geographically defined communities in different ways”. [89] Still, federal elections aren’t about the quality of your local school or the regularity of garbage removal from your street. Most voters, especially in federal elections, just aren’t especially concerned with riding-level matters. To say that expatriates, and only expatriates, ought to be disenfranchised because they aren’t is, once again, to apply an unwarranted double standard.

The existing law already ensures that Canadians voting from abroad cast their ballots in the ridings to which they have the strongest connections. At the same time, it does not require the existence of a very meaningful connection between any voters, including those resident in Canada, and their ridings. The idea that expatriates should be prevented from voting because they lack such a connection is thus a double standard. Moreover, Canadian elections, especially federal ones, aren’t local affairs anyway. For all these reasons, had the government argued that Parliament was entitled to deny expatriates the franchise because of their supposed detachment from the ridings in which their votes would be counted, it would have fared no better than it actually did in Frank.