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,  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),  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″ . 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”.  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”.  This drew on no s 3 precedent whatever, but rather on Libman v Quebec (Attorney General),  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”.  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”.  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. 
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”,  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”.  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”.  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. 
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.” 
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”.  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”.  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.