Toron2(b)

The Supreme Court will hear the appeal of the Toronto election case on March 16, 2021. The case involves the constitutionality of the Better Local Government Act¹2018, S.O. 2018, c. 11 – Bill 5 that redrew the City of Toronto’s electoral districts from 47 wards to 25 wards while the City’s election was ongoing. A variety of applicants sought to challenge the bill. Justice Belobaba granted the initial application, finding some of the provisions of the bill to be unconstitutional. This decision was stayed shortly thereafter by the ONCA (per Hoy ACJO and Sharpe and Trotter JJA) and then reversed 3–2 on appeal (Miller JA with Tulloch and Harvison Young JJA prevailing over MacPherson JA with Nordheimer JA).

The litigation sounded facially implausible: municipalities are famously creatures of the provinces, and have historically been (mis?)managed according to provincial whim. Cities are said to be “bodies corporate”²See City of Toronto Act, s 125(1), and like any corporation can be created, destroyed, or otherwise mucked about with by the appropriate sovereign legislature. Moreover, the Charter‘s election rights guarantees do not extend to municipalities.³“Every citizen of Canada has the right to vote in an election of members of the House of Commons or of a legislative assembly and to be qualified for membership therein.”: Canadian Charter of Rights and Freedoms, s 3 The notion that an “unwritten” constitutional principle forbids meddling in municipal elections seemed similarly implausible: if the municipality could always be destroyed by the province and direct provincial rule imposed, how could it be that a particular form of municipal election was mandatory?

The applicants, cleverly, focused on other sections of the Charter, particularly freedom of expression rights under s 2(b). They urged, and Belobaba J accepted, that freedom of expression of both the Toronto municipal election candidates⁴paras 22–39 and voters⁵paras 40–59 was infringed by redrawing the electoral boundaries amidst the election.

Justice Miller rejected these arguments on appeal, saying that the expression of candidates was not limited, only the effectiveness of that expression at achieving the aim of the candidates.⁶2019 ONCA 732 at paras 41, 44-46 Moreover, in his view, s 2(b) of the Charter cannot include a protection on the right to vote as expressive activity, else s 3 of the Charter (protecting the right to vote in provincial and federal elections) would be surplusage.⁷2019 ONCA 732 at paras 73–76

MacPherson JA, dissenting on appeal, offered instead a chilling effect argument: in his view, if election results can be derailed by legislation, the genuine debate and deliberation that is foundational to elections would be curtailed.⁸2019 ONCA 732 at para 123 Moreover, in this case the legislation had created such a furor that it effectively drowned out the expression of those attempting to engage in a democratic process.⁹2019 ONCA 732 at para 128

None of these opinions is particularly compelling. Justice Miller’s approach is sensible only if there is no connection between the expression of candidates and the legislation; Justice Belobaba’s approach muddies ss 2 and 3; and Justice MacPherson’s injected a wrongful distinction between the period “during” an election and after.

For most legislation, the best way to approach a s 2(b) issue is to identify the consequences the legislation sets out for expression: in Keegstra¹⁰R. v. Keegstra, [1990] 3 SCR 697, this approach would have identified that the legislation set out criminal consequences for hate speech. Such legislation is primarily forward-looking and general and will impose consequences on speech that occurs after the legislation is enacted.

This approach is inappropriate, when, as with the Better Local Government Act, the legislation is alleged to punish speech that predates it. In these circumstances, the court must decide whether expression caused the enactment. If the legislation was intended as a punishment for expression, or a reasonable person would construe it as such,¹¹Note that the Supreme Court jurisprudence on when a chilling effect can be inferred focuses on the effect of a law, not the purpose of the law: see, e.g., R. v. Khawaja, 2012 SCC 69 at paras 79–83. then the legislation probably infringes s 2(b).

Firing an employee can count as such punishment for s 2(b).¹²“In my view, there is little doubt that s. 33 of the Act, which prohibits partisan political expression and activity by public servants under threat of disciplinary action including dismissal from employment, violates the right to freedom of expression in s. 2(b) of the Charter”, Osborne v. Canada (Treasury Board), [1991] 2 SCR 69 per Sopinka J (in majority on this point) Dissolving a job has the same effect, from the perspective of the employee, as that employee being fired, and so it too could count as punishment for the sake of s 2(b). Dissolving a job because of the employee’s speech should be treated the same under s 2(b) as firing the employee. Proving such causation when a job is eliminated may be more difficult than proving causation when an employee is fired — one would not normally expect a government to cut off its nose to spite its face — but if it is proven, then the 2(b) infringement still occurs.¹³Even then, of course, the s 1 analysis may differ.

This framework can be applied straightforwardly to the context of municipal officials. If the mayor of Toronto criticized the government of Ontario, and Ontario in response passed legislation firing the mayor, the mayor’s freedom of expression would clearly be violated. The same goes if Ontario instead dissolved the mayoralty.

The presence of an election does not alter the framework for analysis but may attenuate the chain of causation between expression and the negative consequence. If the mayor-elect criticizes Ontario post-election, and Ontario dissolved the mayoralty in response, the mayor’s freedom of expression is still infringed. If an election is ongoing and the leading candidate criticizes Ontario policy, and in response Ontario cancels the election altogether, the causal chain between their expression and the punishment is more tenuous, because they haven’t won yet. If an election has just begun, and someone new has entered the race and might win, and the election gets cancelled because of the presence of the entrant, the chain is thinner still — but it may exist! In all cases, the relevant question should be, “would the legislature have passed this law, causing this negative consequence for the speaker, but for this expression?”

Applying this framework to the Toronto case, a negative consequence is easy enough to identify: the legislation reduced the number of wards from 47 to 25, thereby eliminating the jobs of 22 councillors. The enactment was also clearly caused by speech, and political speech at that! Ontario’s official justification for the legislation was that debates in Toronto City Council were taking too long¹⁴2018 ONSC 5151 at para 66: literally, there was too much expression, by politicians acting politically. In itself, this suggests that there is a violation of s 2(b) with regard to the reduction in wards.

But there is a second potential infringement of s 2(b) in the Better Local Government Act — one not addressed by lower courts. In addition to the impacts on the City of Toronto, the Better Local Government Act amended s 218.1 of the Municipal Act, 2001. These amendments impacted the government of regional councils for areas outside Toronto. Before the amendments, the heads of seven regional councils were elected directly by their respective populations. The Better Local Government Act cancelled these elections (substituting appointment by the members of the respective councils) for four regions (Muskoka, Niagara, Peel, and York), but maintained the direct elections for three regions (Durham, Halton, and Waterloo). These amendments were widely seen as directed at preventing Patrick Brown, Premier Doug Ford’s predecessor as leader of the Progressive Conservative party, from staging a political comeback as head of council for Peel region.¹⁵Mahoney, J. “Ontario government flip-flopped on cancelling elections for heads of regional governments, records show“, The Globe and Mail (Oct 13, 2019). In other words, the Better Local Government Act appears to have cancelled an election in order to torpedo the career of one of Premier Ford’s political rivals.

The impropriety of such action as a matter of good government is obvious. Whether it infringes s 2(b) is more complicated and depends on why the Ontario legislature in fact promulgated this legislation. If it did so for reasons unrelated to Brown, because it thought that Brown would be a poor regional chair, or out of sheer pique, there may be no infringement. If this was indeed an attempt to derail a political rival¹⁶indeed, since Brown was Premier Ford’s rival in a provincial government, rather than a municipal government, the Act may even violate s 3 of the Charter. A successful political candidate using their newfound powers to punish their former rivals may deter future candidates from running for membership in the provincial legislature. To the victors go the spoils is not exactly the model of electoral competition the Charter intended., or was punishment for things Brown had said, it infringed s 2(b). It may even infringe s 2(b) if the statute was caused by a lack of expression — Brown’s silence. Consider: if Brown had made appropriate obeisances in the direction of Premier Ford, would this legislation have been enacted? If a but-for cause of the legislation (or, for chilling effect purposes, what a reasonable person would understand as a but-for cause of the effect) was that Brown did not flatter Premier Ford sufficiently, the legislation implies forced speech: “kiss up or have your future job destroyed”.

Both of these cases suggest that the constitutional analysis fits uneasily with some of our higher-level norms about appropriate government action. If the legislation was implemented purely, as Belobaba J suggests, out of pique,¹⁷2018 ONSC 5151 at para 70 then there is probably no infringement of s 2(b). Caprice is constitutionally permitted of legislatures, even if it is not permitted of the executive branch of state, as a matter of administrative law.¹⁸This raises the interesting question of whether the legislature could delegate the ability to be capricious. The ban on capricious administrative action arises from courts’ presumption about statutory meaning. Would there be any bar on clear legislative text that authorized an administrative actor to decide a matter arbitrarily or capriciously? If the legislature, however, had a reason and that reason is related to speech, then s 2(b) is apt to be infringed. Irrational bad legislative action may be more constitutionally compliant than rational bad legislative action.

This overall framework also reveals tension between the s 2(b) analysis and the s 1 analysis. Legislative action that is unsupportable under s 1 may not infringe s 2(b); legislative action that infringes s 2(b) will have a basis to back it up according to s 1. This tension is particularly acute with elected municipal government officials. Their expression is inherently political, and so is at the core of freedom of expression¹⁹R. v. Guignard, 2002 SCC 14 at para 20. Inhibiting their political speech is thus a significant infringement of s 2(b). At the same time, however, their political role also means that their speech is not only speech — elected officials directly affect policy outcomes. A strong justification for infringement under s 1 action is that the legislature wishes a different policy outcome.

Despite these tensions, some scenarios are clearly unconstitutional. For example:

1. General legislation that says a municipal election will be re-run if the elected representative is not a member of the Conservative party (forced speech; not justified)
2. Special legislation that cancels an election and replaces the mayor with a provincial appointee after the front-runner publicly disparages Premier Ford (infringement on speech; not justified).
3. Special legislation that cancels an election after a new candidate’s supporters hold a rally outside Queen’s Park to demand that Premier Ford resign (here, the supporters’ speech is punished, not the candidate’s; even so, it is not justified).

By contrast, the next examples are probably constitutional:

4. General legislation that says a municipal election will be re-run if the winner is anyone other than the candidate nominated by the Lieutenant Governor in Council. This would turn an “election” into a referendum on the provincial Cabinet’s choice (not an infringement on speech; or justified if so).
5. Special legislation that cancels an election after the front-runner says they (and council) will shut down the city government until the province provides the city with greater funds (infringement on speech, but justified).
6. Special legislation that replaces the mayor after they (and council) shut down the city government, being silent as to the reason (not itself an infringement on speech).

The more complicated factual situations arise when, say, both 2 and 5 are true. There is then both a plausible legitimate reason (the candidate promises to implement bad policies) for the enactment and a plausible illegitimate one (the candidate criticized the government). The court then must sift through the evidence and determine if the legislature acted for the illegitimate reason on the balance of probabilities.

A particular wrinkle is added to assessing the balance of probabilities by the possibility of a chilling effect on expression. Often, evidence will be indeterminate on its own, and the role of the finder of fact is to collapse the indeterminacy and make a determination. Not so when a chilling effect is in play. A chilling effect can manifest simply when people think that the legislature acted for an inappropriate purpose, even if this is not the court’s finding of fact. Exercising freedom of expression can be deterred by the expectation of a punishment, regardless of the correctness of that expectation.

Of course, an irrational expectation of punishment cannot ground a s 2(b) infringement via a chilling effect. The appropriate standard is probably something somewhat like that for recusal: whether a “reasonable, right-minded and properly informed person” would see the legislation as having been enacted for the improper purpose²⁰See Wewaykum Indian Band v. Canada, 2003 SCC 45 at para 74. If reasonable persons could differ in their assessment of where the balance of probabilities lies according to the evidence, then it is appropriate for a court to consider the chilling effect. Taking this into account would require courts to do something they are oft loath to do: acknowledge both indeterminacy and that despite their supremacy at fact-finding within the legal system, they are not supreme at fact-finding within society as a whole. For individuals in society, a court finding of fact is definitive as regards legal truth, but only probative outside the legal system.

Practically speaking, what does this legal realist prescription entail? Simply a thumb on the scales. If a judge is pretty sure that the legislation was not caused by an inappropriate reason, then the judge should go ahead and find no infringement. Similarly, if the judge is pretty sure that the legislation was caused by the inappropriate speech, the judge should go ahead and find an infringement. But if this is a really tough call that could go either way, then the judge should find an infringement explicitly because of the chilling effect of uncertainty.

Where would this leave the Supreme Court appeal? For the elimination of councillors at the City of Toronto, the s 2(b) violation seems clear: councillors spoke too often, presumably for political purposes, in a way that Premier Ford disliked. Whether this can be upheld under s 1 is debatable. A more minimally impairing measure may have been to impose a limit of 25 speakers per issue on City of Toronto debates (with the other 22 able to make written submissions). This solution, however, may not have accelerated decision-making as much as Ontario intended. Regardless, if Ontario’s provisions have functioned to accelerate decision-making in City Hall, Ontario should be able to adduce evidence of this acceleration by comparing the performance of Toronto City Council before and after the reduction in council size. If Ontario does not do so, the Court should draw the adverse inference that there is no evidence that Ontario’s measure accomplished its ostensible goal.

The Patrick Brown issue (elimination of regional chairs) is probably not properly before the Court, so the Court shouldn’t weigh in on it. If it did, however, the challenge would be to make out the s 2(b) violation. Not having seen any of the evidence, my own prior belief is that this decision was made out of interpersonal and political animosity and that if Patrick Brown had expressed sufficient deference and contrition, that the role he was running for would not have been eliminated. I would thus lean toward finding a s 2(b) violation here as well. But this is a matter of priors, not evidence — on the same evidence (news reports), one might instead draw the inference that nothing Patrick Brown could have said would have rescued him. In that case, there would be no violation. If the Court did find a infringement, the s 1 analysis is straightforward: there is no obvious non-arbitrary reason for eliminating some regional chair elections while maintaining others.