Gas Pump Stickers and Compelled Speech

In my last Slaw post, I discussed Morgan J.’s decision in Canadian Civil Liberties Association v. Attorney General of Ontario (CCLA v. AG Ont.) to grant standing to the Canadian Civil Liberties Association to challenge the government’s requirement that gas station operators post a sticker on each gas pump. The standing decision stands for two major points: the CCLA’s expertise did not have to relate to the selling of gas or regulations governing it, but it was sufficient that it related to constitutional issues; and it had shown its interest by identifying its concerns to the government early, often and in many different ways. This week I address that portion of Morgan J.’s decision dealing with the claim that the sticker infringes freedom of speech.

A quick reminder: in its April 2019 budget, the Ontario government announced it would require gas station operators to post stickers on their gas pumps, which purported to show the increase in the price of gas because of the federal fuel charge. In May 2019, the Legislative Assembly passed the Federal Carbon Tax Transparency Act (FCTTA) imposing the requirement and providing for heavy fines for each day a gas station operator did not comply. Section 5 of the FCTTA identified the content of the sticker and Regulation 275/19 under the FCTTA instructs the operator where to post the sticker on the pump. (See CCLA v. AG Ont., at paras. 1 to 4 for greater detail about the requirements.)

The CCLA sought a declaration that sections 2, 4 and 5 of the FCTTA and Regulation 275/19 constituted an unjustifiable infringement of section 2(b) of the Canadian Charter of Rights and Freedoms, being an instance of compelled speech.

The CCLA argued that the “information” on the sticker was actually political speech that reflected the political agenda of the provincial government in furtherance of its opposition to the federal fuel charge, which was part of the federal government’s environmental plan. Indeed, the CCLA was more specific, perhaps more specific than it needed to be:

It is the CCLA’s position that the key motivation of the governing Progressive Conservatives in Ontario for implementing the Sticker policy, and the specific motivation for its timing in the run-up to the federal election, was to influence the electorate to vote against the incumbent Liberal Party that enacted the FCTTA [sic]. (CCLA v. AG Ont, para. 18)

Although not referring to the electorate’s voting preferences, Morgan J. did conclude that the evidence “demonstrate[s] that the FCTTA should be understood as forming a part of the Ontario government’s engagement with federal partisan politics” (CCLA v. AG Ont, para. 18). He referred to statements by Premier Ford and cabinet ministers (see paras. 19-22).

This assessment of the evidence, which Morgan J. develops as background prior to his section 2(b) analysis, is crucial to the CCLA’s claim that the impact of the government’s requiring the stickers be posted on gas pumps was to draw gas pump operators into the political dispute between the federal and provincial governments (or, put another way, into the attack by the provincial government on the federal fuel charge).

Justice Morgan reviews the case law on freedom of expression, noting three important elements: expression can take “‘an infinite variety of forms'”, which are protected as long as they convey meaning (with very limited exceptions, such as violence), and the freedom “‘encompasses the right not to express views'” (citations omitted; Morgan J.’s emphasis). Minor incursions on freedom of expression may not constitute a contravention, but here “the principle at stake is substantial when seen from the point of view of the retailer, its customers, and of a rights-based society in general” (CCLA v. AG Ont., at para. 44 [emphasis added]).

Justice Morgan bases his analysis on the three questions articulated in McAteer v. Attorney General (Canada), an Ontario Court of Appeal decision (on appeal from a decision of Morgan J.’s). In McAteer, the appellants claimed the requirement of taking the oath to the Queen to become a citizen contravened section 2(b) of the Charter (among other Charter contraventions), as a form of compelled speech. Justice Morgan held that the oath did constitute compelled speech, but was justified under section 1 of the Charter. The Court of Appeal, however, considered there was no contravention, since although the oath “is expressive activity that falls within the ambit of s. 2(b) … the purpose of the oath is not to compel expression; rather, its purpose is to inquire into the would-be citizen’s commitment to our form of government.” Therefore, the incidental effect on freedom of expression “does not warrant constitutional disapprobation of the oath” for several reasons, including that the appellants can “disavow” the oath’s message publicly (McAteer, CA, at paras. 75 and 76). In other words, the oath requirement does not completely deny the appellants’ right to express their views on the matter. (Perhaps not surprisingly given its section 2(b) analysis, the Court of Appeal concluded that if the oath did contravene section 2(b) of the Charter, it was justified under section 1.) In relation to the CCLA’s claim, the relevant questions are as follows:

The first question is whether the activity in which the CCLA is being forced to engage is expression. The second question is whether the purpose of the law is aimed at controlling expression. If it is, a finding of a violation of s. 2(b) is automatic. If the purpose of the law is not to control expression, then in order to establish an infringement of a person’s Charter right, the claimant must show that the law has an adverse effect on expression. In addition, the claimant must demonstrate that the meaning he or she wishes to convey relates to the purposes underlying the guarantee of free expression, such that the law warrants constitutional disapprobation. (CCLA v. AG Ont., at para. 45)

As Justice Morgan says, the answer to the first question is easy: the sticker is a form of expression and because the FCTTA imposes a mandatory requirement under threat of penalty, “the challenged law compels expression” (CCLA v. AG (Ont), at para. 47). (This seems to confound the first and second questions, whether the law compels expression and whether the law compels expression, respectively.)

The second question, however, shifts the analysis somewhat. According to McAteer, the second question is actually whether the purpose of the law controls expression. The CCLA’s claim is that the law compels expression. Justice Morgan agrees that the the sticker is expression, but the second question is not whether “the purpose of the law is aimed at compelling expression”, but “whether the purpose of the law is aimed at controlling expression” (CCLA v. AG Ont., at para. 48 [emphasis in original]. This reflects the Court of Appeal’s wording in McAteer.

The question here is the difference between “compelling” and “controlling”. In adopting the Court of Appeal’s approach in McAteer, Morgan J. has also adopted the confusion that inheres in thinking about the issue this way. The term “compelled speech” has been coined to differentiate it from “denied speech”. Initially, contraventions of section 2(b) were “assumed” to be the result of governments’ efforts to prevent people from expressing their views; however, it became clear that requiring people to say things they didn’t want to say may be equally invasive of their Charter right to free expression. Hence the term “compelled speech”.

In answering the third question (“does the law adversely effect [sic] expression, and is this effect related to the purposes of the freedom of expression guarantee?” (CCLA v. AG Ont., at para. 53)), Morgan J. distinguishes McAteer and the case before him as follows: “there is one crucial difference between the compelled speech in McAteer and the compelled speech under the FCTTA, and that is that the speech at issue in McAteer was of a nature that fostered the rule of law whereas the speech at issue here arguably countermands the rule of law.” (CCLA v. AG Ont., at para. 54)

The sticker uses incorrect language (“Federal Carbon Tax” instead of “Fuel Charge”), which, since the Court of Appeal had already determined the fuel charge is a valid regulatory provision and not a tax, “would be perfectly acceptable in political advertising or in a politician’s speech, but it is an intentional use of ‘spin’ that reveals the advocacy rather than informational thrust of the message.” (CCLA v. AG Ont., at para. 60)

The inaccurate language and incomplete information (there is no reference to the rebate for provinces and tax relief for individuals) is intended to deliver the message that “the incumbent party in Ontario has better policy ideas than the incumbent party in Ottawa. That is a perfectly acceptable message in a political campaign, but it is not the one that the government has purported to enact.” (CCLA v. AG Ont., at para. 64) The Ontario legislature has thus undermined the purposes underlying freedom of expression.

In considering whether the provisions are justifiable under section 1 of the Charter, Morgan J. considers them

in light of what they expressly convey, what they omit, the context in which they were enacted, and the statements in the legislature of the government members that introduced the measures. Sticking strictly to the legislative language is only reliable where the context supports the language used. Without putting too fine a point on it, scholars of statutory interpretation have observed that statutes are enacted by legislators/politicians, and that “politicians sometimes misrepresent their actual policy preferences” (citation omitted) (CCLA v. AG Ont., at para. 66)

Since the objective of the legislation and regulation is to promote Ontario’s political position vis-a-vis the federal government’s legislation and policy (“partisan advantage-seeking”), it does not constitute “a pressing and substantial government objective”. But even if they passed the first stage of the section 1 analysis, “The salutary benefit of the information conveyed by requiring a Sticker on every gas tank would be far outweighed by the deleterious effect of using legislative and regulatory power for partisan ends.” (CCLA v. AG Ont., at para. 70) As for the last stage, “It would be an insurmountable hurdle for the Ontario to require a sticker that uses explicitly partisan rather than ostensibly neutral language, and to then establish that the benefit of its bona fide public policy choice outweighed the detriment of its self-interested, partisan use of governmental power.” (CCLA v. AG Ont., at para. 78)

The McAteer test can be usefully compared to how Deschamps J. framed the test in Canadian Broadcasting:

…[T]o determine whether an expressive activity is protected by the Charter, we must answer three questions: (1) Does the activity in question have expressive content, thereby bringing it, prima facie, within the scope of s. 2(b) protection? (2) Is the activity excluded from that protection as a result of either the location or the method of expression? (3) If the activity is protected, does an infringement of the protected right result from either the purpose or the effect of the government action? (Canadian Broadcasting, at para. 38 [citations omitted])

While the courts, including the Supreme Court of Canada, have employed the concept of “control” in describing the government’s action, it does not have to form part of the test for a section 2(b) violation. The McAteer approach addresses a different question that might more properly be considered under section 1: the extent to which the government’s action affects the opportunity for expression by both the applicant or other people. As Morgan J. explains (but in referring to the second question), “[t]he question posed by the McAteer analysis … is a different one: is the law designed not so much to compel a specific message but to control any message the gasoline retailer might otherwise express?” (CCLA v. AG Ont, at para. 48). “Control” comes close to saying that the denial or compulsion covers the field, including not only those whose speech is directly affected and who is bringing the claim, but others, as well. Of course, it is possible that this can happen, but in the usual case, the denial or compulsion will be partial, either in content, location or in whom it affects.

McAteer relies on Irwin Toy for the analysis relating to controlling speech. It can be argued, however, that the use of “control” in Irwin Toy is not meant to add a step or to the analysis or to change the way the challenge is framed, but rather simply to indicate that the government has done something that restricts the applicant’s expression. In Irwin Toy, the way government “controlled” Irwin Toy’s speech was to limit what company could do with advertising. Thus Irwin Toy’s ability to advertise in a way that would attract children was denied.

The term “control” is better seen as a synonym for either “deny” or “compel” than being the way the test is framed in McAteer and adopted by Morgan J. Using it to frame the test means that it is necessary to take into account whether the applicant is able to use other means to deliver the message that has been controlled. This begins to place parameters around the broad understanding of “freedom of expression” under section 2(b). In McAteer itself, for example, the Court of Appeal explained that the applicants could express their views about the citizenship oath and therefore the requirement to take the oath did not have a chilling effect on expression: “The opportunity to disavow the message is relevant to the determination of whether a chilling effect will occur.” (McAteer, at para. 77).

As the Court of Appeal in McAteer pointed out, McLachlin C.J.C., speaking for the Court, relied in part on the concept of a “chilling effect” in Khawaja to conclude that the terrorism offences under which Khawaja was charged did not contravene section 2(b). In Khawaja, it was the “motive clause” in the legislation that Khawaja argued would have a chilling effect because it defined terrorism in part by reference to the motive’s being a political, religious or ideological purpose, objective or cause. In referring to “a chilling effect”, McLachlin J. seems to be thinking about whether other people will be concerned about expressing political, religious or ideological views similar to those expressed by persons caught by the anti-terrorism provisions (for various reasons, she concluded the motive clause did not have a chilling effect).

As Morgan J. pointed out, the gas station operators in CCLA v. AG Ont. could have created their own stickers with information missing from the government’s sticker (such as that the GGPPA not only imposes a fuel surcharge, but also provides for a fuel rebate) their own point of view and posted them on the gas pumps. Accordingly, “the law is designed to compel expression but not to control it in the constitutionally understood sense” and thus there is not an automatic violation of section 2(b)” (CCLA v. AG Ont., at para. 52).

In earlier cases, this assessment revolved around whether it was the government’s purpose to deny or compel speech; if so, there was an automatic infringement of section 2(b). If the impugned provision’s effect was to deny or compel expression, the analysis at this stage became more nuanced. As “[h]aving found that the plaintiff’s activity does fall within the scope of guaranteed free expression, it must next be determined whether the purpose or effect of the impugned governmental action was to control attempts to convey meaning through that activity.” (Irwin Toy, per Dickson CJC and Lamer and Wilson JJ). This does not mean complete control necesarily. I note the term “control” appears throughout this analysis, applying whether the purpose or effect of the provisions are at issue; it does not seem to constitute in itself a step in the test, but appears to be used as a synonym for “restricts”. And it can refer to controlling access by others or by the applicant.

Is there another way to think about control that helps to simplify the analysis? Government controls speech by denying someone the right to speak or take particular actions or by requiring someone to speak or take a particular action when they do not with to do so. This should satisfy the second question and in this context, deny or compel alone is sufficient. However, how much control, over the applicant and over others — whether there is a chilling effect — may well be relevant to determining purpose or effect or how carefully the government has fashioned the restriction.

The blatant and some might think crude comments to explain the sticker played yeoman service in Morgan J.’s analysis: it formed part of his background discussion in which he appears to accept the CCLA’s argument that the legislation and regulation “are intrinsically wed to partisan politics”, which was then relevant to his granting standing because it may have discouraged gas station operators from participating in the case; it also relevant to the first stage of the Charter analysis, whether the law has infringed freedom of expression (because it is manifested in the sticker’s content); and it is relevant to Morgan J.’s section 1 analysis because it contributes to the government’s failure at the first stage of the test under section 1 (the incorrect and incomplete content make the government, in Morgan J.’s words at para. 68, “an unreliable narrator”).

There is a sense here that the government has not acted ethically, although Morgan J. does not explicitly say so, or has somehow failed in its proper role by importing partisan sniping and conduct into the more austere realm of constitutional analysis. It may be this that Morgan J. means when he says that one of the interests at stake is that of “a rights-based society in general”. A rights-based society deserves better.

It is worth asking whether the gas pump sticker would be unconstitutional if the government had merely sent it out to gas station operators to post on their gas pumps, even with the penalties that do apply, had it not shot itself in the foot with blatant ministerial comments about its purpose. Would it then be vulnerable because it provided incomplete (and false) information? Would it be as easy to conclude that its purpose was to be blatantly partisan or would it merely appear sloppy? Are people protected under section 2(b) if they are required to post a sign that is incorrect if it cannot be shown that its flaws are intended and not merely careless? The real offence here is that private individuals were being required to assist the government in delivering a partisan message, whether they wanted to do so or not. The chilling effect derived not from whether they or commentators generally felt they could not criticize either the sticker or the provincial government’s position but from the mandatory nature of the legislation coupled with the potentially (and disproportionately) heavy penalties that were intended to advance the government’s political agenda. The subtext of the decision in CCLA v. AG Ont. is that the government should be wary of abusing its power — or it should be more careful about how it does so.