Unstuck

Ontario’s Superior Court strikes down the anti-carbon tax-sticker law, but still doesn’t get freedom of expression

Last year, I wrote about Ontario’s Federal Carbon Tax Transparency Act, 2019 (the “Act”) ― both about the disgraceful way in which it became law and about its unconstitutional speech compulsion, which I argued should not even be considered as a potentially justified limitation of the freedom of expression under section 1 of the Canadian Charter of Rights and Freedoms, because it was tantamount to the imposition of an official ideology. The constitutionality of the Act was in fact challenged by the Canadian Civil Liberties Association and, last week, in CCLA v Ontario (Attorney General), 2020 ONSC 4838, the Superior Court of Ontario struck it down.

At first glance, this is a welcome development for the freedom from compelled speech. Not only is the compulsion invalidated, but Justice Morgan’s approach might seem to bear some resemblance to the one I had proposed: in effect, he denies the government the chance to justify the Act under section 1. But look at Justice Morgan’s reasons more closely, and they turn out to be very narrow. Indeed, they could be used to support significant speech compulsions in the future.

This is not altogether surprising. Justice Morgan was constrained by the Court of Appeal’s reasoning in McAteer v. Canada (Attorney General), 2014 ONCA 578, 121 OR (3d) 1, upholding the constitutionality of the requirement that applicants for the Canadian citizenship swear an oath to bear “true allegiance” to the “Queen of Canada”, which I have described as a “parade of horribles“. And indeed it was none other than Justice Morgan who had written the first instance decision in that case. While it wasn’t quite as bad as that of the Court of Appeal, it did not evince much understanding of the harms of compelled speech either.


The Act required all gas stations to display a prescribed sticker alerting customers to the amount of the “federal carbon tax” levied on the gas they were purchasing. The evidence adduced by the CCLA showed that it was meant as a not-so-subtle intervention in the 2019 federal election campaign, in which the Ontario government supported the anti-carbon-tax position of the federal Conservatives and opposed the pro-carbon-tax Liberals. This partisan dynamic is a key factor in Justice Morgan’s reasoning.

Before getting to the substantive issues, Justice Morgan must address the Attorney General’s objection to the CCLA’s standing to challenge the Act. As it turns out, the CCLA has tried to enlist actual gas stations as plaintiffs or co-plaintiffs, but none would come forward. Justice Morgan explains that “retailers, with a view to market forces rather than to politics and constitutional law, have been loath to participate in this case” due to its political valence. [40] But the record to which Justice Morgan alludes suggests that this is not quite accurate: politics, in the shape of a fear of regulatory retaliation, seems to have been a motivating factor too. Be that as it may, Justice Morgan grants the CCLA public interest standing to pursue the case.

He must next decide whether the sticker requirement limits the freedom of expression protected by section 2(b) of the Charter. To this end, he applies the test set out in the Court of Appeal’s McAteer decision:

The first question is whether the activity in which the plaintiff 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. (McAteer, [69])

Justice Morgan finds that the sticker is indeed a form of expression. Yet in his view its purpose is not to control expression. In particular, he takes the view that “it would be difficult for the government to control expression by compelling certain messages … but not restricting others”. [50] Objectors remain free “to disavow” [52] the message they are compelled to voice, for example by posting disclaimers; hence their expression is not “controlled”. However, it is adversely impacted by the Act.

The key point for Justice Morgan is that, unlike the citizenship oath in McAteer, the sticker does not promote democracy and the Rule of Law. Indeed, it does not even serve to truthfully inform. Justice Morgan attaches some importance to the sticker’s use of the “carbon tax” nomenclature, which in his view is at odds with the Court of Appeal for Ontario’s opinion, in Reference re Greenhouse Gas Pollution Pricing Act, 2019 ONCA 544, 146 OR (3d) 65, that the policy at issue is not a “tax” within the meaning of the Constitution Act, 1867. Moreover, the sticker says nothing of the ways the money levied as carbon tax is distributed, in part to taxpayers, in part to provinces. As a result, it is a form of “spin”. [60] By requiring it, “the government is not so much explaining a policy [as] making a partisan argument”. [63] And “[b]y using law for partisan ends, the Ontario legislature has enacted a measure that runs counter to, rather in furtherance of, the purposes underlying freedom of expression”. [65]

This limitation of the freedom of expression is not justified under section 1 of the Charter. Indeed, unusually, Justice Morgan finds that the Act lacks a pressing and substantial purpose ― the first, and normally very low, hurdle a statute must meet to be upheld under section 1: “While truly informing the public about the components that make up the cost of gasoline would be a pressing and substantial purpose, promoting the Ontario governing party over the federal governing party is not.” [69] The Act is purely partisan rather than a real “policy choice”. [69] Justice Morgan goes through the other steps of the justified limitation analysis by way of an obiter, but it all comes down to his concern with partisanship. The Act is invalid.


Right outcome, but the reasoning is another matter entirely. Justice Morgan’s approach is illogical and conflicts with the Supreme Court’s precedents, notably inthat it collapses the two stages of the Charter analysis that the Supreme Court has always sought to keep distinct: first, the question of whether a right is being limited; second, that of whether the limitation is justified.

First, to say, as Justice Morgan does, that one’s expression is not controlled because one can disavow something one has been coerced to say is perverse. The fact that one is forced into disavowals shows sufficiently that what one is saying is not what one chooses to say.

The political context that Justice Morgan’s reasons depict highlights this problem. As he explains, it appears that gas station owners would rather keep quiet and sit out the political conflict about the carbon tax. This is their right ― the obverse of the freedom of speech is the freedom to stay silent. If they are forced into disavowals and denials, the gas stations will inevitably be taking sides in the political conflict they are trying to avoid ― if anything, this will be much more obvious than if they merely comply with the Act and display the required stickers. Of course, such a response is not what the Ontario legislature envisioned, but it would be caused entirely by the Act, and so it is absurd to deny that the Act amounts to a form of control of the gas stations’ expression.

Fortunately, the Supreme Court’s precedents mandate no such twisted inquiry. Ostensibly the most important freedom of expression case (I have argued here that it is only “leading from behind”), and the source of the “control” language used in McAteer and by Justice Morgan is Irwin Toy Ltd v Quebec (Attorney General), [1989] 1 SCR 927. The distinction between legislation that has control of expression as purpose and that which doesn’t is described as follows in the joint opinion of Chief Justice Dickson and Justices Lamer and Wilson:

If the government’s purpose is to restrict the content of expression by singling out particular meanings that are not to be conveyed, it necessarily limits the guarantee of free expression. If the government’s purpose is to restrict a form of expression in order to control access by others to the meaning being conveyed or to control the ability of the one conveying the meaning to do so, it also limits the guarantee. On the other hand, where the government aims to control only the physical consequences of certain human activity, regardless of the meaning being conveyed, its purpose is not to control expression. (974)

Applied to speech compulsions rather than censorship, this means that any legislation that “singles out particular meanings” that must be communicated, or forces an audience to listen to a communication, necessarily has control of expression as its purpose. Such legislation limits (or, as the Supreme Court often says, prima facie infringes) the freedom of expression. There is no need to consider effects, let alone to ask the purely subjective question of whether they are worthy of “constitutional disapprobation”.

This inquiry into effects and “disapprobation” in effect forces claimants to show that the law which compels their speech is not justified, and more specifically that it pursues an end worthy of judicial condemnation. The success of such an argument in this case should not blind us to the fact that this is a high hurdle. As noted above, this approach collapses the usual section 1 test of whether a limitation on a right is justified into the threshold inquiry of whether a right is limited in the first place, and it means that the claimant rather than the government bears the burden of proof. It follows that Justice Morgan’s streamlined approach to the section 1 analysis is rather less supportive of freedom of expression than one might think. The important work is already done by the time he gets there, as he has, in effect, found that the Act is unjustifiable. Had he not so found, he would have upheld it without ever getting to section 1, just as the Court of Appeal upheld the citizenship oath in McAteer.

Last but not least, Justice Morgan’s emphasis on partisanship as the fundamental problem with the Act is also misguided. For one thing, as tempting as it might be to say that partisanship can never be a sufficient justification for restricting Charter rights, the Supreme Court has in the past upheld laws that protect political incumbents from criticism, notably in Harper v Canada (Attorney General), 2004 SCC 33, [2004] 1 SCR 827. I think the Supreme Court was wrong to disregard the partisan valence of that legislation, but this shows that it will often be difficult to disentangle partisanship from other, specious considerations. Indeed, Justice Morgan himself suggests that a statute that is “a hybrid of policy and partisanship” would deserve to be treated as fully legitimate.

More importantly, Justice Morgan’s understanding of partisanship is very narrow. It does not encompass the promotion of a state ideology that rises, if perhaps only slightly, above the “horse race” version of partisan politics. He has nothing but sympathy for governmental “protection and promotion of Canada’s national and legal culture” [58] by means of forcing those who did not agree with this culture to voice loyalty to it. Needless to say, there is a political dimension to a “national and legal culture”, especially when this culture is coercively imposed by the state, even though Justice Morgan is oblivious to this. To him the distinctions between partisanship and high principle appear obvious. To the rest of us living in 2020, they are anything but.

Consider an obvious example: the late and unlamented “statement of principles” requirement that the Law Society of Ontario tried to impose on its members. Certainly its supporters argued in terms promoting a certain high-minded vision of social and legal culture (indeed they spoke of a “culture shift”). But then again, as we now know, there is a bitter partisan division over the issue within the ranks of the Law Society’s membership. So how would Justice Morgan approach the question of the constitutionality of the requirement? And would his approach be different now than it would have been before the partisan cleavage was revealed by the success of the StopSOP campaign in the 2019 Bencher election? Whatever we might think of the “statement of principles” or its opponents (of whom I was one), or of compelled speech more broadly, I hope we can agree that this is not a reasonable way of addressing such an important issue.


Of course it is a good thing that the Act is no more, and that the Ontario government, if it wants to continue its anti-carbon-tax propaganda campaign, will have to do it by itself, rather than by means of conscripting third parties. I have argued here that such ideological conscription is wrong when it serves to supposedly advance some rights-protecting agenda. It is no less wrong, obviously, when its aim has to do with fiscal and environmental policy. Governments have plenty of resources at their command. If they want to propagandize, they have no need to get unwilling individuals to do it for them.

Yet, the state of the law on compelled speech, and indeed on freedom of expression more generally, in Ontario at least, is cause for concern. It’s not just that few restrictions on freedom of expression are ever struck down. More importantly, the courts fail to understand what free speech means, and why it matters. Justice Morgan’s reasons for striking down the Act illustrate these failures just as much as his and the Court of Appeal’s earlier reasons for upholding the citizenship oath did.

Ministers of Truth

A proposal to criminalize epidemic-related “misinformation” is dangerous

The CBC’s Elizabeth Thompson reports on a rather startling development: the federal government is, apparently, giving serious thought to introducing censorship to discussions of the present plague. More specifically, there is talk of “legislation to make it an offence to knowingly spread misinformation that could harm people”, based on a member of the UK House of Commons proposal “for laws to punish those responsible for spreading dangerous misinformation online about the COVID-19 pandemic”. At least some of the opposition seem keen, Ms. Thompson quoting an NDP Member of Parliament as claiming that “Extraordinary times require extraordinary measures and it is about protecting the public”, and reassuring us, I suppose, that “[t]his is not a question of freedom of speech”.

Actually, it is very much a question of freedom of speech. The Supreme Court invalidated an earlier prohibition on the speading of “false” news in R v Zundel, [1992] 2 SCR 731, and for good reason. Such prohibitions mean that government telling us what we are and what we are not allowed to say. Say something the government deems, in the words of the same honourable gentleman, to “interfere with the efforts of our frontline medical workers”, and suffer punishment. This is a limitation of the freedom of speech on any plausible definition of that concept, and for a Member of Parliament to pretend otherwise is not only an illustration of the politicians’ habitual mendacity but, more specifically, a rather ironic way of getting the public used to the idea of meting out punishment for statements that fail to live up to a standard of truth.

It is far from clear just what these restrictions are meant to accomplish. The CBC report quotes a spokesperson from the Communications Security Establishment, an intelligence agency, as warning about “cybercriminals and fraudsters” who “encourage victims to visit fake web sites, open email attachments and click on text message links” that purport to provide health information. But fraud, for example, is already a crime; there is no need for “extraordinary measures” to prohibit it, or for broadly defined bans on “misinformation”. The report also says that “Health Canada … is sending compliance letters to companies it finds making false or questionable claims about COVID-19”. It is not quite clear what sort of compliance is in question here, but presumably ― or at least hopefully ― it’s compliance with existing laws, perhaps ones having to do with advertising, or specifically advertising of health products. If so, then why is more legislation necessary?

For his part, the NDP MP tells, darkly, of “troll bot farms, state operators or … conspiracy theorist cranks who seem to get their kicks out of creating havoc”. State actors with troll bot farms at their disposal are unlikely to be deterred by Canadian legislation. At most, then, it will be targeting conspiracy theorists… and giving them more ammunition for believing the government is hiding things. Is there any evidence at all, actually, that “conspiracy theorist cranks” ― especially ones within the reach of Canadian laws, and not the one domiciled at 1600 Pennsylvania Ave., Washington, DC ― are having a real effect on Canada’s response to the plague?

And on the other side of the scales, there will be real costs to this proposed legislation. Even if it includes the mens rea requirements of knowledge, wilfulness, and malice ― which, if applied, would result in good faith conspiracist cranks being off the hook ― the law is likely to produce chilling effects. Worse, attempts to enforce it, even if they do not ultimately lead to convictions, will target the politically unpopular, or simply those who happen for one reason or another, to incur the displeasure of police services and prosecutors. As concerning as recent stories of overzealous enforcement of “social distancing” regulations are, the problem is much more longstanding one. Readers may remember me blogging about a makeup artist prosecuted for gory videos involving no actual gore or violence whatsoever and Québec blogger who ― stupidly, to be sure ― mused about a mass shooting in the legislature, about the man who had to go all the way to the Ontario Court of Appeal to quash a municipality’s attempt to prosecute him for a solitary, non-violent protest in the town square, and about the author and publisher of a novel prosecuted for a brief and not remotely titillating description of the rape of a child. And the provisions invoked in these cases are all well-known, and not directed at dealing with a crisis. There is every chance that an emergency anti-disinformation law will result in harsh and arbitrary prosecutions. Even if the accused are ultimately acquitted, they will have undergone considerable stress and expense in the meantime. And, again ― for what?

Even in the short term, the harm of a law against plague-related “disinformation” is likely to outweigh what little good it might do. But the real damage it will do will occur in the medium and long term, as it becomes a template for widespread criminalization of statements deemed to be contrary to this or that state policy. The British MP whose ideas are inspiring the Canadian proposals is apparently drawing his own inspiration from “Germany’s laws governing online hate speech or France’s legislation countering disinformation during election campaigns”. And the report itself notes that the federal

government set up an elaborate system to watch out for attempts to disrupt last year’s federal election through disinformation, including a committee that brought together several departments and a special group chaired by the clerk of the Privy Council to sound the alarm.

Once the plague is over, it will be all too tempting to declare something else the next great public emergency, and to repurpose, instead of abolishing, the censorship mechanisms that allow government to silence those who question or undermine its response ― even if stupidly.

If there is there one thing we’ve learned from events of barely a year ago, it’s that clerks of the Privy Council are not always imbued with a great respect for constitutional propriety, or immune to the temptation to shill for their political masters. I would not trust one of them with the job of a Minister of Truth. Nor would I trust the public health authorities, which themselves at times seem quite confused about what the truth is. Indeed, this confusion only serves to underlie the fact that a government that is entitled to impose the truth on its subjects ― who can no longer be counted as citizens ― is also a government that is empowered to lie to them. No one, after, is allowed, and at length able, to tell the difference. The Canadian government needs to reverse course before it becomes a government of this sort.

Day Nine: Leonid Sirota

The Roads Not Taken

Sometimes, as other contributors to the symposium have discussed, dissenting opinions chart the law’s future course. But at other times, they are only signposts for alternative paths which the law passes by, perhaps for the better. And sometimes, they point to the lost straight road, from which the law tragically deviates, never to return. The three dissents below belong to this last category.


1. Justice Beetz in Slaight Communications v Davidson, [1989] 1 SCR 1038

Slaight was an unjust dismissal case, in which a labour arbitrator sided with the former employee. The issue at the Supreme Court was the arbitrator could, consistently with the Charter, require the former employer to provide the employee with a recommendation letter bearing the employer’s signature but actually entirely dictated by the arbitrator, and further to refrain from saying anything else about the former employee. The majority held that he could. After all, there was a power imbalance between employer and employee that needed to be rectified, and anyway the employer was only required to state true facts, as established by the arbitrator.

Justice Beetz saw things differently. To force a person to state “facts in which, rightly or wrongly, he may not believe” is tantamount making him “tell a lie”. The outcome of an official fact-finding process cannot be equated with an objective, all-purpose truth, let alone be elevated into a dogma everyone must believe in. The state has no more authority to make a person proclaim what it, but not he, believes to be true facts than to make him proclaim what it, but not he, believes to be true opinions. Such an order “is totalitarian in nature and can never be justified under s. 1 of the Charter. It does not differ, essentially, from the command given to Galileo by the Inquisition to abjure the cosmology of Copernicus.”

Justice Beetz also rejected the arbitrator’s order that the former employer not say anything other than what the arbitrator required about the former employee. He pointed out that “one should view with extreme suspicion an administrative order or even a judicial order which has the effect of preventing the litigants from commenting upon and even criticizing the rulings of the deciding board or court”. Finally, while condemning the former employer, Justice Beetz pointed out that “under the Charter, freedom of opinion and freedom of expression are guaranteed to ‘everyone’, employers and employees alike, irrespective of their labour practices and of their bargaining power.”

All these points are important, and Canadian law is the worse for not having taken them more seriously. Most disturbingly, of course, we have seen in recent years recurring attempts to impose official dogma on dissenting individuals, whether by the Law Society of Ontario or by the governments of Canada and Ontario. But we also now have an asymmetrical Charter jurisprudence, notably in the realm of freedom of association, against which Justice Beetz correctly warned. And, while fortunately we have not seen attempts to stifle criticism of the judiciary or the administrative state by law, too many Canadian lawyers are intolerant of critiques of their judicial heroes.

2. Justice McLachlin (as she then was) in R v Keegstra, [1990] 3 SCR 697

Before she became, allegedly, the “Conscience-in-Chief” of Canada, or at least of the Central Canadian establishment, and a Chief Justice somewhat notorious for strong-arming colleagues into consensus, Justice McLachlin, as she once was, authored a number of important dissents. Famously, the one in Rodriguez v British Columbia (Attorney-General), [1993] 3 SCR 519 eventually, in effect, became Supreme Court’s unanimous position. The one in Keegstra did not. Even Chief Justice McLachlin, as she became, eventually resiled from it. That’s too bad.

In Keegstra, the Supreme Court considered the constitutionality of the Criminal Code‘s proscription of hate speech. The four-judge majority upheld it as a reasonable limit on the freedom of expression. Justice McLachlin wrote for three (on the freedom of expression issue) dissenters. Her opinion is, perhaps, a little fastidious, and contains little in the way of memorable language, but it is thoughtful and deserves to be considered even by those who do not ultimately agree with her. Indeed, having argued the substantive case against the criminalization of hate speech elsewhere on this blog (and Emmett Macfarlane having discussed them in his contribution to this symposium), it is the more general or procedural points that I would like to highlight here.

For one thing, Justice McLachlin was fundamentally skeptical of content-based regulation of speech, and much sympathetic to the American approach, the views all such regulation with great suspicion. For another, Justice McLachlin firmly rejected the attempt to equate hate speech with violence. Violence, she stressed, involved the use of physical force, not words, even hurtful words. Furthermore, Justice McLachlin refused to read down the Charter‘s protection of freedom of expression in the name of equality: “it seems a misapplication of Charter values to … limit the scope of that individual guarantee [of freedom of expression] with an argument based on s. 15, which is also aimed at circumscribing the power of the state”. Compare this to the use of “Charter values” to impose egalitarianism on private actors and eviscerate religious freedom in Law Society of British Columbia v Trinity Western University, 2018 SCC 32, [2018] 2 SCR 293!

Last but not least, consider Justice McLachlin’s insistence on the need for evidence to justify limitations on the freedom of expression. While acknowledging the appropriateness of some deference to the government on this issue, Justice McLachlin nevertheless wrote that, in order to avoid trivializing the justification of limitations on rights, “in cases … where it appears that the legislation not only may fail to achieve its goal but may have a contrary effect, the Court is justified in finding that the rational connection between the measure and the objective is absent”. Good intentions are not enough ― nor is the sort of ill-informed speculation, camouflaged as “common sense”, that has all too often sufficed in subsequent Supreme Court decisions.

Had just one vote gone the other way, and this opinion become the law, our constitution may well have been in much better shape than it is, far beyond the narrow issue of hate speech. As things stand, Keegstra has to count as one of the more significant missed opportunities in the Charter‘s history.

3. Justice Moldaver in Reference re Supreme Court Act, ss 5 and 6, 2014 SCC 21, [2014] 1 SCR 433, a.k.a. l’Affaire Nadon

In l’Affaire Nadon the Supreme Court was asked to opine on the eligibility of the judges of federal courts for appointment to the Supreme Court itself, and especially that of judges of the federal courts from Québec for appointment to one the Supreme Court’s Québec seats. It was, as readers will recall, a very high-profile and controversial case (more on which in a forthcoming book by Michael Plaxton and Carissima Mathen). The sort of case, in other words, in which the Supreme Court not infrequently issues unanimous opinions “by the court”. But Justice Moldaver’s dissent prevented the majority from giving itself this ultimate institutional imprimatur.

The majority held that, while judges of the federal courts were, as former lawyers, eligible for non-Québec seats on the Supreme Court, only current lawyers or current judges of the Québec’s superior courts could take one of the Québec seats. In doing so, the majority relied heavily on the idea that judges from Québec had to be not only experts in the civil law, but also representatives of Québec’s “social values”. This, they could not do without being current, not merely former, judges of Québec’s courts or members of the Québec bar.

For his part, Justice Moldaver dissected each of the majority’s arguments, and found them empty. In particular, as a matter of text, the two provisions governing eligibility for appointment ― the general one requiring judges to be or to “ha[ve] been” judges or lawyers of at least 10 years’ experience, and the specific one providing that Québec judges are to be chosen “from among the judges of the Court of Appeal or of the Superior Court of the Province of Quebec or from among the advocates of that Province” ― are “inextricably linked”. If the 10-year rule applies to Québec seats, as the majority accepted, so must the eligibility of former lawyers.

As for purpose, Justice Moldaver rejected the majority’s claim that the eligibility criteria had anything to do with the representation of Québec’s alleged “social values”. Indeed, “[i]mporting social values — 140 years later — is unsupported by the text and history of the [Supreme Court] Act”. The majority’s interpretation leads to the absurd result that judges not only of the federal courts, but also of Québec’s provincial court, are ineligible for appointment, while a lawyer who has done no more than pay his fees to the Québec bar while not engaging with the law at all could be appointed; so could a former judge who rejoined the Québec bar for a single day. While Parliament might have chosen such absurd criteria for eligibility and said so, “when interpreting a statute to determine what the relevant criteria are — i.e. what Parliament intended them to be — absurd results are to be avoided”.

As I have said here before, the majority opinion was not only wrong but pernicious; in particular, its linchpin, the concept of “social values”, was just self-important twaddle. Justice Moldaver deserves credit for exposing its vacuity. Rumour has it that he did it at some cost to himself. His fortitude, then, is to be commended as much as his legal acumen.


Honourable mentions: Justices Brown and Côté in Law Society of British Columbia v. Trinity Western University, 2018 SCC 32, [2018] 2 S.C.R. 293, which I described here as “probably the best opinion to come out of the Supreme Court in a long while”, and Justices Martland and Ritchie in Re: Resolution to amend the Constitution, [1981] 1 SCR 753, a.k.a the Patriation Reference, which I plan on discussing further in a post on unwritten constitutional principles in a not-too-distant future.

Day Three: Emmett Macfarlane

Among the panoply of difficult constitutional decisions rendered by the Supreme Court of Canada, there are many occasions when the majority of justices provide reasoning that can only be described as less than compelling (some might simply say ‘wrong’). The virtues of dissenting reasons – which, even on a highly consensual court like the Supreme Court, appear in roughly one-quarter of all cases – are multiple, and include presenting a counterpoint that might sharpen the overall decision, identifying weaknesses in the majority’s reasons, and, perhaps most importantly, providing a potential foundation for a future iteration of the Court to overturn itself (indeed, this has happened in cases involving assisted dying and labour rights.) 

An invitation to identify three favourite dissents poses a considerable challenge given the long list of candidates, but I’ve managed to settle on the following:

  • Dissenting opinion in R. v. Keegstra (1990), by Justice McLachlin (as she then was). 

The Keegstra case involved a Charter of Rights challenge to the criminal law against unlawfully promoting hatred. The majority upheld the law as a reasonable limit of freedom of expression. They did so in part on the basis that hate speech “is of limited importance when measured against free expression values … the state should not be the sole arbiter of truth, but neither should we overplay the view that rationality will overcome all falsehoods in the unregulated marketplace of ideas.” Moreover, hate speech subverts the democratic process by denying dignity to at least some segments of the community. The majority is dismissive of efforts to “prove a causative link between a specific statement and hatred of an identifiable group” and even states that requiring such proof of direct harm “would severely debilitate” Parliament’s objectives. Instead, it is enough that there is a risk of harm.

McLachlin’s dissent acknowledges the intuitive kinds of harm that hate speech can generate, particularly the pain and indignity it can inflict upon its targets. Yet she rightly questions the effectiveness of criminalizing hate speech. Indeed, the law is rarely enforced in Canada precisely because it does not capture that vast majority of hateful utterances. McLachlin also notes that hatred is notoriously broad, and that identifying it requires reliance on vague or subjective understandings. Importantly, this had already resulted in dramatic state overreach. She points to instances where copies of Salmon Rushdie’s The Satanic Verses were stopped by border authorities in a misguided effort to enforce the criminal provision. In another incident, arrests were made when pamphlets were distributed that happened to include the words “Yankee Go Home.”

The Keegstra dissent is a principled defence of free expression and the dangers of permitting state line-drawing on a vague basis like the promotion of hatred. McLachlin’s dissent correctly highlights the lack of evidence that hate speech laws mitigate hateful expression, the very real risk of state overreach, and the chilling effect such laws might induce. It is a shame that, when offered a chance to revisit the issue of hate speech in the statutory human rights context years later in Saskatchewan (Human Rights Commission) v, Whatcott, McLachlin essentially disregarded her own important points of caution.

The Chaoulli case involved, at its core, a fundamental principle of the design of the health care system – equity, specifically access to health care regardless of ability to pay – and whether a provision designed to protect it, the prohibition on the purchase of private medical insurance, violated the right to life, liberty and security of the person under section 7 of the Charter. Had there been clear evidence that the provision at stake in the case contributed to waitlists and delays in access to health care, this might have been a straightforward decision. But regardless of whether one supports, as a matter of policy, a greater role for private options in health care, everyone should be concerned about the majority’s capacity to properly assess the evidence at stake in the case.

In a remarkably frank and punchy dissent, Justices Binnie and LeBel excoriate their colleagues for their overconfidence and questionable assumptions in deciding that the law ought to be invalidated. Noting that their colleagues contend the failure to provide “public health care of a reasonable standard within a reasonable time” violated rights, the dissenters ask:

What, then, are constitutionally required “reasonable health services”?  What is treatment “within a reasonable time”?  What are the benchmarks?  How short a waiting list is short enough?  How many MRIs does the Constitution require?  The majority does not tell us.  The majority lays down no manageable constitutional standard.  The public cannot know, nor can judges or governments know, how much health care is “reasonable” enough … It is to be hoped that we will know it when we see it.

The dissent rightly criticizes the majority for a lack of deference to finding of facts at the trial level, for disregarding the majority of experts, and for failing to pay heed to comparative evidence that waitlists exist in countries with private options. In a particularly noteworthy passage for a Supreme Court of Canada opinion of any kind, the dissent notes bluntly that the “resolution of such a complex fact-laden policy debate does not fit easily within the institutional competence or procedures of courts of law.” Moreover, they note that a “legislative policy is not ‘arbitrary’ just because we may disagree with it.” If only this message was one Canadian justices heeded more often.

The Remuneration reference is one of the most dramatic cases of judicial overreach in Canadian history. In it, the majority of the Court mandated “independent compensation commissions” for judges based on the “unwritten principle” of judicial independence (grounded in the preamble to the Constitution Act, 1867 of “a Constitution similar in Principle to that of the United Kingdom” and an analysis of section 11(d) of the Charter, a plain reading of which comes nowhere close to imagining the requirements invented by the majority).

Justice La Forest’s partial dissent stands as the lone voice of reason on a Court wildly stretching and misapplying the concept of judicial independence. He describes the majority’s approach as “a partial usurpation of the provinces’ power to set the salaries of inferior court judges” under the Constitution Act, 1867. That the reference involved “an issue on which judges can hardly be seen to be indifferent, especially as it concerns their own remuneration” was not lost on him either. La Forest criticizes the majority for its view that the constitutional preamble is a source for limiting the power of legislatures to interfere with judicial independence. Indeed, the idea that the British Constitution imposes such limits on Parliament is ahistorical nonsense.

La Forest also correctly notes that judicial review is “politically legitimate only insofar as it involves the interpretation of an authoritative constitutional instrument. … That legitimacy is imperiled, however, when courts attempt to limit the power of legislatures without recourse to express textual authority.” It is unreasonable, in La Forest’s view, to assume changes in judicial salaries or discussions between the two branches of government about salaries impair judicial independence.

Honourable mentions:

The dissent in Daviault (1994), against a defence of extreme intoxication for offense of general intent like sexual assault.

The dissent in Saskatchewan Federation of Labour (2015), against constitutionalizing the right to strike.

The dissent in Reference re Supreme Court Act (2014), against a cherry-picked connection between the general eligibility requirements for Supreme Court justices and those for judges from Quebec.

The dissent in R. v. N.S. (2012), against the notion that requiring a sexual assault complainant to remove her niqab when testifying at trial protects the right to a fair trial.

The dissent in Sauvé (2002), in favour of deference to Parliament’s legitimate moral and philosophical objectives in denying the right to vote to those currently in prison for having committed serious crimes.

 

Day Two: Kerri A. Froc

The Power of Saying No

University of New Brunswick

The ability to reject traditional reasoning, to say “no”, is a central part of feminist critique and practice. Student groups introduced the “no means no” campaign into popular consciousness over two decades ago to emphasize the importance of sexual consent. While it lost purchase because of its seeming implicit burden imposed on women to communicate non-consent, the original idea behind it was to shift cultural values. Women’s “no” could no longer be devalued as meaningless, or a challenge to be overcome, worse yet, as a disingenuous way of saying “yes”. 

Feminists often have to say “no” a lot, in terms of positively asserting that they reject inequitable, conventional understandings and refuse to go along. In a patriarchal culture, that becomes read as “sex negative”, as overly sensitive, or as biased (as the Chief Justice of the Quebec Court of Appeal recently discovered). While saying “no” is often powerful and sometimes a moral imperative, it wears on you.  As Ahmed says:

[A] no can still be dismissed as impertinent in the sense of rudely bold or boldly rude and can be judged as an act of political vandalism. So many refusals are dismissed in these terms; you might be free to say no but your no is heard as destructive; hearings have consequences (becoming a killjoy is a consequence)… For feminism: no is political labour.

So, in the dissents I want to talk about, I celebrate the refusal to “go along” in favour of what might be professionally risky for the judge or simply a great deal of effort wasted or ignored.  They represent ways of thinking that deserve another look.

Justice Frank Iacobucci in Little Sisters Book and Art Emporium v Canada (Commissioner of Customs and Revenue), [2000] 2 SCR 1120

In Little Sisters, federal customs officials, under the auspices of holding back material they deemed “obscene” under the Customs Act, targeted a lesbian bookstore for discriminatory treatment. While the majority found that there was discrimination in application, this could not be attributed to the Act itself, as “Parliament is entitled to proceed on the basis that its enactments ‘will be applied constitutionally’ by the public service.” This was notwithstanding that customs officers were ill-trained to identify obscene material, and that the process for challenging improper decision-making was lengthy and cumbersome. The multipart, legalistic Butler regime to determine obscenity was deemed sufficient to guide officers – the problem was not that the Act but the individuals applying its rules.

By contrast, Iaccobucci refused this characterization – he saw the problems as systemic and “baked into” the regime established by the Act, leading to lack of training, turnover of officers, lack of procedural fairness for importers, as well as “superficial and context-insensitive” review of materials. He found accordingly that the Act “practically invites” violations of s.2(b) freedom of expression.  The framework needed to be completely rethought from the perspective of expressive rights. 

His seems to be an eminently appropriate approach where a regime is so flawed that it can be reasonably anticipated that its operation will very likely result in rights violations.  In the same way, manufacturers cannot avoid tort liability completely where they make products with built-in design flaws, notwithstanding that harm could be avoided if those using them did so perfectly (rather than like typical human beings). Surely, fundamental constitutional rights demand at least as much protection (especially as those affected cannot avoid the state’s “product”). History proved Iacobucci J. right – Little Sisters continued to be targeted notwithstanding government promises in the Supreme Court appeal that it had improved the administration of the Act.  Nevertheless, I have tried in vain to find any court decisions where his dissent on this point has been taken up and applied.

Justice Claire L’Heureux-Dubé in Thibaudeau v Canada, [1995] 2 SCR 627

To borrow a phrase, Thibaudeau is a terrible, horrible, no good, very bad equality decision, and the split amongst the judges is telling: the two female judges on the Court wrote separate dissents, with the male majority rejecting the section 15 claim. At issue was the treatment of child support under the Income Tax Act: it was taxable in the hands of custodial parents (98% of which were women) and a tax deduction for payors. Both justices analyzed the income tax regime through a gender lens, one that accounted for the realities of women raising children alone. The male judges relied on abstractions and legal fictions. 

The majority and concurring decisions found that there was no negative distinction, because in most cases there was a net tax benefit to the “family unit” (which no longer existed).  They thereby deemed egalitarian distribution of benefits to exist in families (even defunct ones!). This would notionally occur in separated families through “gross-ups” of child support to take into account tax consequences. If this version of “trickle down economics” did not occur and women had less money for their children, this inequality was “peculiar to specific cases” – the fault of individual judges or fathers not passing along tax benefits – and not the Income Tax Act. The justices refused to consider the unequal impact on custodial parents in their own right. 

In Thibaudeau, L’Heureux-Dubé J. gestures to the doctrine of coverture in underscoring how using the couple as the unit of analysis for adverse effects of the taxation rules obscures inequality.  She recognized the Act as the source of detrimental treatment because its “default” is that the benefit accrues completely to the non-custodial spouse and the detriment to the custodial spouse. The onus is on the custodial spouse to “wage an unremitting and costly battle, both emotionally and in the family law system,” if the family law system was to remedy the inequality completely through the gross-up mechanism.  Not only did this require judges to perform repeated calculations perfectly, it also did not consider the practical realities of separated family life with custodial parents – women – having less money for legal fees and needing to avoid antagonizing non-custodial spouses. Despite her reasons not carrying the day in court, Parliament was persuaded: child support became non-deductible, non-taxable in 1997.

This case is emblematic of the justice’s emphatic “no” to an analysis of a woman’s Charter case that is degendered and abstracted to the point of absurdity.  Constance Backhouse in her oeuvre, Claire L’Heureux-Dubé: A Life, documents the cost L’Heureux-Dubé J. pays for her rejections, including a fractious relationship with Justice (later Chief Justice) Antonio Lamer and a public, gendered attack by an appellate court judge following a sexual assault appeal popularly referred to as the “no means no” case, R v Ewanchuk, [1999] 1 SCR 330.

Her call to recognize the detrimental impact of default regimes that confer benefits to the more advantaged spouse would take nearly 20 years to be finally be recognized, in Quebec v A, 2013 SCC 5, [2013] 1 SCR 61.  Even then, the majority voted to justify provincial family law legislation excluding common law spouses under section 1 because it prioritized autonomy and “choice” of couples.   An approach that fully attends to conditions of subordination in which such “default” legislation operates has therefore yet to be fully embraced.

Justice Bertha Wilson in R v Morgentaler, [1988] 1 SCR 30

I wrestled with selecting the last “dissent”: do I adhere to the letter of these blog posts (dissents = a decision that is directly contrary to the majority on outcome) or the spirit (dissents = minority opinions that should have carried the day but didn’t)?  There are several other decisions in which I agree with the dissenters on outcome, but their reasons are not completely compelling. In the end, I decided to keep with the spirit and discuss an opinion that technically is a concurrence. 

At the time of Morgentaler (1988), Madam Justice Wilson was the only woman in a court that was not hospitable to women members, which makes her opinion even more remarkable for her refusal to “go along.” Relatively well known by now is that Justice Wilson departed from the majority by ruling that not only did the Criminal Code therapeutic abortion committee regime violate women’s security of the person due to its imposition of psychological trauma and unnecessary physical risk, any restrictions on abortion violated women’s right to liberty. She redefined liberty to include the right to make fundamental decision over one’s own life free from state interference, which included the decision as to whether to carry a pregnancy to term. In doing so, she created a more inclusive and objective conception of liberty over that gendered male. She remarked that the history of human rights had been “the history of men struggling to assert their dignity and common humanity against an overbearing state apparatus”, to the exclusion of “women’s needs and aspirations are only now being translated into protected rights”. Her conception of liberty came to inform majority decisions of the Court in cases like Blencoe v British Columbia (Human Rights Commission), [2000] 2 SCR 307.

Less cited, however, is how she found that the regime violated the principles of fundamental justice. She noted that Justice Lamer referred to other rights in sections 8-14 in interpreting fundamental justice in BC Motor Vehicles, thus leading her to surmise that the concept means not only procedural fairness but also consistency with other Charter rights and freedoms. Accordingly, “a deprivation of the s. 7  right which has the effect of infringing a right guaranteed elsewhere in the Charter  cannot be in accordance with the principles of fundamental justice.” The abortion regime also infringed women’s freedom of conscience given that the state sought to override women’s own moral decision-making. Consideration of other rights violations as a breach of fundamental justice is profound – it recognizes that the Charter as a whole is an embodiment of what is just. In an article called “Constitutional Coalescence”, I argued this does not necessarily mean that one does, e.g. a mini-s.2(a) or s.15 analysis within section 7, but that an interpreter views the former rights through a different (potentially wider) lens, one that goes beyond a hyper-individualized and procedurally-based notion of justice to one that considers systemic structures of subordination. This is in stark contrast to other cases in which the Court has been at pains to keep rights conceptually separate and has declined to consider all rights in multiple rights claims. This led to what I refer to in my earlier work as a “watertight compartments” approach to the Charter leading to complete rejection of claims involving multiple rights. Despite its potential enrichment to our understanding of Charter rights, Wilson J’s innovation has not explicitly been taken up by other judges.


As Carissima Mathen has written in relation to equality,  a divided decision “that is the result of failure to reach agreement on ‘deep’ issues is preferable to one that, as the price of unanimity, remains ‘shallow.’”  The dissents that I have highlighted reflect the potential depth of dissenting decisions, and into which I hope future justices will mine for their wealth. 

Upcoming Canadian Talks

Save the dates!

In a couple of weeks, I will be hopping on to a 13-hour transpacific flight and heading to Canada to give a series of talks. Here are the dates and topics. I don’t have all the details about the exact time and location yet, so if you are based at or near one of the host institutions, keep an eye out ― or get in touch with me or my hosts closer to the day.

  • September 26, University of Victoria, Faculty of Law (Runnymede Society): “The Road to Serfdom, 75 Years On”. I take it that this will be inaugural Runnymede event at UVic, and I am very honoured to be part of it.
  • September 30, Université de Sherbrooke, Faculty of Law: « Route de la Servitude: fermée pour travaux (de démolition)… depuis 75 ans ». This will be the French version of the UVic talk; I’m afraid I’m a bit puzzled by the title, but I didn’t to choose it.
  • October 2, University of Toronto, Faculty of Law (Runnymede Society): “An Election Is No Time to Discuss Serious Issues. Really?” This will be discussion of the regulation of civil society participation in election campaigns, which has been much in the news in recent weeks.
  • October 4, University of Waterloo (Freedom of Expression in Canada Workshop): “A Conscience- and Integrity-Based Approach to Compelled Speech”. The workshop is being organized by Emmett Macfarlane, who has just told it is full… but there is apparently a waitlist. My paper builds, of course, on what I have had to say about things like the citizenship oath, the Law Society of Ontario’s “statement of principles”, and Ontario’s anti-carbon-tax stickers.
  • October 9, Université du Québec à Montréal, Département des sciences juridiques: « Les élections sont-elles une occasion de se taire? ». This will be the French version of the Toronto talk, with a discussion of the Québec legislation thrown in.
  • October 11-12, Ottawa (Workshop on the Royal Prerogative): “The Royal Prerogative in New Zealand”. This is the first meeting of a group put together by Philippe Lagassé to carry out a SSHRC-funded research project on the prerogative in Canada, the UK, Australia, and New Zealand. Professor Lagassé also tells me the workshop is “pretty much full”. Are you seeing a theme here? Yep, I’ve managed to get myself invited to really cool workshops.
  • October 16, McGill University, Faculty of Law (Runnymede Society): a discussion with Paul Daly on administrative law. If the Supreme Court co-operates, we will, of course, discuss the Vavilov and Bell/NFL cases, in which the Court may, or may not, completely change the Canadian law of judicial review. If the decisions are not released, it will be a more general conversation. Either way, I am looking forward to
  • October 18, Université de Montréal (Symposium of the Journal of Commonwealth Law): “Unholy Trinity: The Failure of Administrative Constitutionalism in Canada”. I will be presenting a paper arguing that the Supreme Court’s disgraceful decision in Law Society of British Columbia v Trinity Western University, 2018 SCC 32, [2018] 2 SCR 293 and the companion Ontario case illustrate the problems that plague “administrative constitutionalism” ― the view that administrative decision-makers’ decisions bearing on constitutional rights are entitled to judicial deference.

I am grateful to the people who have invited me and/or organized these events. (A special shout-out to my co-blogger and president of the Runnymede Society, Mark Mancini!) If you are able to make it to one (or more) of the talks, please say hello. It is always a pleasure to meet some of my readers in person. See you soon!

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.

Do Not Pass Section 1: Go Directly to Invalidity

Some infringements on rights are never acceptable in a free and democratic society, including requirements to state facts one doesn’t believe in

In my last post, I argued that Ontario’s recently-enacted and not-yet-in-force Federal Carbon Tax Transparency Act, 2019, which requires gas stations to display stickers purporting to inform their clients of the cost of the federal carbon tax, is likely unconstitutional, as well as morally wrong. The requirement obviously compels the owners of gas stations to engage in speech from which they would otherwise have abstained, and so limits their right to freedom of expression protected by section 2(b) of the Canadian Charter of Rights and Freedoms. In my last post, I followed the orthodox approach to ascertaining whether this limitation was justified and therefore constitutional, which consists in applying a proportionality analysis along the lines first set out in R v Oakes, [1986] 1 SCR 103. But, as I indicated there, I actually think that this approach is not right for this case. Here, I explain why.


Pursuant to section 1 of the Charter, the rights the Charter protects can be “subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society”. Proportionality analysis is not an end in itself or an explicit requirement of the constitutional text. It is only a means to the end of ascertaining whether a given limitation on rights is “demonstrably justified”. (Indeed, one may well argue that the proportionality analysis is a bad means to that end; one would not be wrong; but it is much easier to poke holes in proportionality analysis than to come up with a convincing all-purpose alternative.) Proportionality analysis is inherently case-by-case. It focuses a court’s attention on the reasons for and the effects of particular statutory provisions or administrative decisions, applied to the particular circumstances detailed by the persons whose rights are allegedly infringed.

But it should be possible to say that certain limitations of rights are such as to be categorically impermissible in a free and democratic society, regardless of particular circumstances. With limitations of this sort, proportionality analysis is unnecessary; indeed, it only serves to obfuscate their inherent unacceptability. I can see no bar in the text of section 1 of the Charter to taking this position. In his article on the history of section 1, Adam Dodek notes that a number of groups that took part in the proceedings of the Special Joint Committee of the Senate and the House of Commons on the Constitution pushed for some rights (equality rights were a popular candidate, but not the only one) to be excluded from the scope of the application of section 1 altogether. Obviously, this was not done, but I don’t think that this rejection entails that of a more fine-grained approach. In other words, while the history may suggest that no provision of the Charter is absolutely immune from limitation, at least as a textual matter, it does not follow that any and all limitations conceivable are, potentially, justifiable in a free and democratic society.

Indeed, I think that it does not follow that a categorical bar on justifying limitations of certain rights, introduced in the process of constitutional construction, is foreclosed by section 1, even in light of the history described by Dean Dodek. The idea that section 1 had to apply to every right guaranteed by the Charter was put to the Supreme Court in Attorney General) v Quebec Association of Protestant School Boards, [1984] 2 SCR 66, but the Court accepted it “for the sake of discussion only and without deciding the point”. In any case, this is an issue for another day.

And there are precedents, in early Charter cases, for applying the approach that I am considering. Protestant School Boards is one. There, the Supreme Court observed that limits on rights, within the meaning of section 1 of the Charter,

cannot be exceptions to the rights and freedoms guaranteed by the Charter nor amount to amendments of the Charter. An Act of Parliament or of a legislature which, for example, purported to impose the beliefs of a State religion would be in direct conflict with s. 2(a) of the Charter, which guarantees freedom of conscience and religion, and would have to be ruled of no force or effect without the necessity of even considering whether such legislation could be legitimized by s. 1. (88)

But the best known precedent is R v Big M Drug Mart Ltd, [1985] 1 SCR 295. There, Justice Dickson (as he then was), wrote that

it should be noted that not every government interest or policy objective is entitled to s. 1 consideration. Principles will have to be developed for recognizing which government objectives are of sufficient importance to warrant overriding a constitutionally protected right or freedom. Once a sufficiently significant government interest is recognized then it must be decided if the means chosen to achieve this interest are reasonable―a form of proportionality test. (352)

Justice Dickson went on to reject the government’s attempt to justify the Lord’s Day Act, which imposed the Christian holy day as a mandatory day of rest for most Canadian workers. He found that

[t]he characterization of the purpose of the Act as one which compels religious observance renders it unnecessary to decide the question of whether s. 1 could validate such legislation whose purpose was otherwise or whether the evidence would be sufficient to discharge the onus upon the appellant to demonstrate the justification advanced. (353)

However, the proportionality analysis foreshadowed in Big M and sketched out by now-Chief Justice Dickson in Oakes quickly took over Charter cases, and the possibility that some limitations of Charter rights could never be justified, regardless of the circumstances and the evidence the government brings in their support has been a road not taken by Canadian constitutional law in the last 35 years.


I think that this unfortunate. The Oakes-based proportionality analysis, at least as it has developed, focuses on one part of section 1: the “demonstrably justified” requirement. But it has little to say about other parts of section 1: the “democratic society” qualifier, and the notion of “limits” on, as opposed to exceptions to or denials of rights. Perhaps it didn’t have to be this way. In Oakes itself, Chief Justice Dickson wrote referred to this phrase as “the final standard of justification for limits on rights and freedoms” (136) and offered an explanation of what they referred to:

the values and principles essential to a free and democratic society …  I believe embody, to name but a few, respect for the inherent dignity of the human person, commitment to social justice and equality, accommodation of a wide variety of beliefs, respect for cultural and group identity, and faith in social and political institutions which enhance the participation of individuals and groups in society. (136)

One might quarrel with this list, of course ― I am not a fan “social justice” as an inherent component of democracy, for instance ― or, at least, expect it to be refined as cases develop. More fundamentally, one might quarrel with the way Chief Justice Dickson proffers this catalogue of values, as the product of his own meditation on freedom and democracy. An originalist, for example, might want to ask what the words “free and democratic society” meant to the public at the time of the Charter‘s enactment, and not simply how a judge ― even a thoughtful and distinguished judge writing mere years after the Charter came into force ― understood them. But, however that may be, the idea that limitations of rights must be justifiable not just in the abstract, but in a particular kind of society, namely a free and democratic one, was there in Oakes ― and has (like certain other aspects of that decision) fallen by the wayside since.

To repeat, I would like to recover this idea and, more specifically, to argue that there are some limits on rights that are never acceptable in free and democratic societies. Protestant School Boards offers and Big M applies one example: it is not acceptable, in free and democratic society, to impose a state religion. One might imagine a specious proportionality-based defence of the Lord’s Day Act: it serves the objective of social cohesion and public affirmation of a national religion, in a way that could not be achieved by less restrictive means, and after all it is but a small imposition ― dissentients are not forcibly dragged to divine service ― in comparison with purported benefits. A sufficiently deferential court might even, conceivably, swallow this. But we don’t need ask whether it would. The alleged benefits of the Lord’s Day Act are not something a government is entitled to pursue in a free and democratic society.

I tentatively think that a similar argument can be made with respect to many speech compulsions. In particular, I think that a free and democratic society is necessarily one in which there is no official ideology prescribed by the state that citizens are required to parrot. I suspect that the idea would have been familiar at the time of the Charter‘s framing, during the Cold War. Thus the rejection of official ideologies may well be part of the original meaning of the phrase “free and democratic society”, although I don’t know enough to be confident. But even if it cannot be read into section 1 as a matter of interpretation, I think that it has to be as a matter of construction ― the process of elaboration of legal doctrine implementing constitutional text. Just like a free and democratic society has no state religion, as the Supreme Court confirmed in Big M, it must have no set of secular beliefs mandatory for citizens. Perhaps having an official ideology would be convenient or useful; perhaps it would foster equality, or social cohesion, or prosperity. This doesn’t matter. Free and democratic societies don’t do official ideology ― just like they don’t do official history, official economic theory or, I would add, official science. (Official, of course, in the sense of mandatory for citizens; the state itself can, and indeed must to some extent, commit to specific views on many of these issues.)

Now, some cases of compelled speech cannot rightly be described as or assimilated to attempts to impose a state ideology. This is, in particular, the case of mandatory disclosure of information that is in the possession of the person subject to the compulsion ― whether in the shape of nutritional information that is required to be printed on food packing or that of data about trust accounts or self-study hours that lawyers are made to provide on their annual reports. Mandatory requirements to use a particular language for certain communications are in this category too. For these, and perhaps other, cases of compelled speech, the proportionality framework, with its case-by-case scrutiny of the tailoring of means to ends and weighing of costs and benefits is appropriate (assuming, that is, that it is appropriate for anything).

At the other extreme are cases like the Law Society of Ontario’s requirement that lawyers “promote equality, diversity, or inclusion”. This is a clear case where the government ― through the entity to which it has delegate coercive regulatory powers over the legal profession ― attempts to force people to embrace a particular set of values or beliefs and express their having done so. One can argue ― along with Dwight Newman ― that this is also an infringement of the freedom of thought (protected by the same provision of the Charter as freedom of expression, section 2(b)). One can also argue, as I have done here, that this is an infringement of the freedom of conscience. But of course this is also (and neither Professor Newman nor I deny this) a limitation of the freedom of expression ― and, I think, a limitation of a sort that cannot be justified in a free and democratic society, no matter how well-intentioned (which it is) or effective (which it isn’t).

The ant-carbon-tax stickers are something of an intermediate case. They ostensibly communicate information, and at least make no pretense about this information coming from the person coerced into transmitting it rather than the government. To that extent, they are less offensive, and less like an official ideology, than the Law Society of Ontario’s demands. However, it is arguable that stickers present incomplete information, and do so tendentiously. Not everyone, to say the least, would regard the message conveyed by the stickers as something that they could, in good faith, transmit. This is more than just a matter of preference. Perhaps the sellers of junk food would rather not tell people the number of calories their product contains; but their integrity is not threatened when they are made to do so. By contrast, when a person is made to communicate something that he or she does not, in good faith, believe, the stakes are higher, and the analogy to official ideology much closer. At the risk of being a bit dramatic, making Winston Smith love Big Brother was only the end point. The start was making him say that 2+2=5.

As Justice Beetz insisted in his dissenting opinion in Slaight Communications Inc v Davidson, [1989] 1 SCR 1038, to accept that it is permissible to order a person to tell the truth “beg[s] the essential question: what is the truth?” (1060) Some authority may think that it has established the facts, but one “cannot be forced to acknowledge and state them as the truth apart from his belief in their veracity. If he states these facts … as ordered, but does not believe them to be true, he does not tell the truth, he tells a lie.” (1061) Justice Beetz went on to add that

to order the affirmation of facts, apart from belief in their veracity by the person who is ordered to affirm them, constitutes a … serious violation of the freedoms of opinion and expression … [S]uch a violation is totalitarian in nature and can never be justified under s. 1 of the Charter. It does not differ, essentially, from the command given to Galileo by the Inquisition to abjure the cosmology of Copernicus. (1061)

Of course, Justice Beetz’s opinion was a dissenting one. All I can say is that I see nothing in Chief Justice Dickson’s majority opinion that addresses his colleague’s cogent arguments. Given the extent to which the Supreme Court has been willing to revisit its prior cases ― and to do so with much less justification than there would be to revisit Slaight on this point ― I feel no particular compunction in urging that Justice Beetz’s opinion should be followed, and that compelled statements of facts that the person required to make them believes, in good faith, to false or simply misleading should be treated like compelled statements of opinion and compelled professions of value. They are categorically unjustifiable in a free and democratic society.


The Charter‘s reference to “a free and democratic society” is not a mere description. As the Supreme Court held early on, it is the “final standard” against which purported limitations on the rights the Charter secures must be measured. It is true that rights must sometimes be limited, even in a free and democratic society. But the Charter exists because of a recognition by its framers ― and by their constituents ― that legislative majorities are apt to disregard rights, and to seek to limit them for the sake of convenience, or out of ignorance or even spite or hatred. Some limitations may appear defensible in principle but, on closer examination, are not supported by evidence, go too far, or do more harm than good. But others are incompatible with free and democratic societies as a matter of principle. It is unnecessary to scrutinize their tailoring to their purpose, or weigh up their effects. The Charter bars them categorically.

The imposition of official beliefs, or the requirement to express beliefs, is the sort of thing that simply must not happen in a free and democratic society; it is incompatible with freedom and democracy. This includes religious beliefs, as the Supreme Court has held. But political beliefs, or even beliefs about truth, should not be treated any differently. Canadian governments have no right to impose them, and the courts should peremptorily reject them.

Sticking It to the Feds

Why Ontario’s anti-carbon-tax stickers are likely unconstitutional, and certainly immoral

It is time, finally, for me to get back to the carbon-tax stickers. Last month, I was distracted from writing this post by my horror at the abusive, indecent way Ontario’s Federal Carbon Tax Transparency Act, 2019, was set to become law. It has now been enacted (though not yet come into force) and, though my disgust at the process of its enactment is unabated, I turn to its substance. The Act is, I believe, unconstitutional. It is also, quite apart from constitutional issues, morally objectionable in its own right, and doubly so coming from a government that ― cynically ― positioned itself as a champion of free speech.

The Act is simple enough. Its only substantive provision requires every “person who is licensed … to operate a retail outlet at which gasoline is sold at a gasoline pump and put into the fuel tanks of motor vehicles” to

obtain from the Minister [of Energy, Northern Development and Mines] copies of the prescribed notice with respect to the price of gasoline sold in Ontario; and … ensure the notice … is affixed to each gasoline pump at the retail outlet in such manner as may be prescribed.

There are also provisions for inspections and fines. The “prescribed notice” is, of course, the notorious sticker.


This is a requirement that all those (individuals or corporations) engaged in a particular trade communicate a message prescribed by the government. In simpler terms, an instance of compelled speech. Under a sane freedom of expression jurisprudence, this must, of course, be regarded as a limitation on the freedom of expression. Whether Ontario currently enjoys the blessings of a sane freedom of expression jurisprudence is open to some doubt, given the holding of the province’s Court of Appeal in McAteer v Canada (Attorney General), 2014 ONCA 578 that the requirement that applicants for Canadian citizenship swear a prescribed oath is not a limitation of the freedom of expression protected by section 2(b) of the Canadian Charter of Rights and Freedoms. However, I think it is best to assume that, on this point at least, McAteer was an aberrant decision that can be disregarded. The carbon tax sticker requirement ought to be held to be a limitation on the section 2(b) right.

There are two paths that one can take from here. The orthodox one, which I shall take in this post, consists in asking whether this limitation is one that can be demonstrably justified in a free and democratic society, and so authorized by section 1 of the Charter. To be justified in a free and democratic society, a limitation on a right protected by the Charter must meet the following criteria, as recently summarized by the Court of Appeal in Christian Medical and Dental Society of Canada v College of Physicians and Surgeons of Ontario, 2019 ONCA 393:

the objective of the impugned measure must be of sufficient importance to warrant overriding a constitutionally protected right or freedom;

… the means chosen must be reasonable and demonstrably justified – this is a “form of proportionality test” which will vary in the circumstances, but requires a balancing of the interests of society with the interests of individuals and groups and has three components:

(i) the measure must be rationally connected to the objective – i.e., carefully designed to achieve the objective and not arbitrary, unfair or based on irrational considerations;

(ii) the means chosen should impair the Charter right or freedom as little as possible; and

(iii) there must be proportionality between the salutary and deleterious effects of the measure. [98]

In my next post, I will suggest that this approach is not appropriate for cases that involve certain types of compelled speech, including this one. More specifically, I will argue that the proportionality analysis can be bypassed in the case of many speech compulsions, which are never appropriate in a free and democratic society. That said, an attempt to follow the ordinary proportionality framework here does the Ontario government few favours.


It is difficult to see what the important objective that warrants the imposition of the stickers is. If one is in charitable mood, one might say that the legislature is really trying to provide transparency about the effects of a public policy that affects Ontario’s consumers. (Less charitably, and perhaps more plausibly, one might say that the the objective here is to score some political points off of the feds.) I don’t think that this an inherently bad thing for a government to do, as Patricia Hughes comes close to saying in a post at Slaw. (Dr. Hughes faults the stickers for “not advanc[ing] an alternative approach to fighting climate change” and, instead, “undermin[ing] an approach that has been widely accepted as a positive response to … greenhouse emissions”. I’m not sure why this would be constitutionally problematic. A bad choice of priorities, perhaps, but this is a debate that courts should probably stay out of.) But even if transparency of this sort is desirable, is it, as the Court of Appeal put it, “of sufficient importance to warrant overriding a constitutionally protected right or freedom”? In theory at least, it should be possible to conceive of objectives that, while desirable, are not worth abridging rights for, and I would argue that this is one of them. Perfectly transparent public policy might be a supererogatory good in a free and democratic society, but not one to be pursued at the expense of such a society’s fundamental commitments, which is what constitutional rights are supposed to be. To be sure, the courts generally tend to be very deferential to legislatures at this stage, but even this deference might, just, have its limits ― and if so, this would be pretty good case to discover them.

Now, assuming that the objective of fostering transparency about the effects of public policy does warrant limitation of rights ― a big assumption, as just explained ― I think it has to follow that the sticker requirement is rationally connected to the objective. The issue at this stage isn’t whether it is a particularly good way of achieving the legislature’s purpose, but whether it’s not an arbitrary one. This is a low bar to clear. Dr. Hughes writes that “[t]here is no rational connection between the message of the stickers … and opposition to the carbon tax because they fail to provide all the information”. Perhaps so, but I don’t think that failure to provide complete information is really a rationality issue, or that courts should be in the business of evaluating the content of a government message to assess its completeness.

What the courts can and should do, however, is to find that conscription of gas stations to communicate the government’s message about the effects of the carbon tax is not the least restrictive means of accomplishing whatever transparency-promoting aims the government might have. Being able to help itself to both the bully pulpit and the public purse to further its public-relations strategies, the government can do without conscripting private parties to carry its water. I am no fan of the Supreme Court’s decision in Canada (Attorney General) v JTI-Macdonald Corp, 2007 SCC 30, [2007] 2 SCR 610, which upheld, among other things, a requirement that tobacco manufacturers display government-mandated health warnings on 50% of their packaging, but it is surely arguable that the warnings regarding the health consequences of a particular product really do need to be displayed on that product, and not elsewhere, to be optimally effective. An argument along these lines is not so easy to make in support of Ontario’s carbon tax stickers. That said, a lot will depend on the level of deference the courts accord the legislature. One suspects, however that a legislature at odds with a carbon tax will be given less deference than one trying to discourage smoking. (This is, I am afraid, not to the Canadian courts’ credit.)

Finally, I think the courts can and should find that the benefits of the stickers, if there are any, are not worth the imposition on those who have no desire to display them. But here too, much depends on the level of scrutiny courts are willing to apply. In JTI-Macdonald, the sum total of Chief Justice McLachlin’s reasoning on this point was “proportionality of effects is established. The benefits flowing from the larger warnings are clear. The detriments to the manufacturers’ expressive interest in creative packaging are small.” [139] If a student could not come up with something more than this conclusory assertion, I would flunk her. But, quod licit Jovi, etc. In any case, here again, the courts’ biases are likely to be less favourable to the legislature, and chances are that the sticker mandate will, in fact be scrutinized as it ought to be.


Whatever doubt there might be about the legal side of the issue (and I don’t think there should be too much), the immorality of the carbon tax sticker requirement is clear. As noted above, the Ontario government has virtually unlimited resources to make its views of the federal carbon tax known. These views, at this point, aren’t exactly a secret, anyhow. But if the government wants to instruct its trained seals MPPs to end their speeches with anti-carbon-tax perorations in the style of Cato the Elder, it can. If it wants to put up giant anti-carbon-tax posters on every town square in the province, it can. If it wants to buy advertising slots from willing newspapers or radio and television stations, it can. Instead of doing the work of communicating its position itself ― and paying to do so, if necessary ―, the government conscripts unwilling private citizens and companies to serve as its bullhorn.

This is beyond its rightful powers, not only on a libertarian or classical liberal conception of the government’s proper powers but also, I think, on either a “progressive” or a conservative one. It is, indeed, little more than than naked abuse of power. The Ontario government makes people do things just because it thinks it can. I have argued here against the view the governments can in effect conscript private individuals to advance their constitutional agendas, or that the Law Society of Ontario can force lawyers to act as advocates on its behalf by “promot[ing] equality, diversity, and inclusion”. The same principles apply to a government’s attempt to communicate its views of public policy. This is something that the government can and must do on its own. If it can force citizens to do that, it can force them to do anything.

Notice, by the way, that this is not just an objection to government mandates to communicate misleading or incomplete information, or messages that undermine policy designed to deal with climate change or whatever other problem. The objection to government conscription of individuals to speak on its behalf is neutral and general. It applies to “progressive” causes, as well as to populist ones. Some means are wrong regardless of the rightness of the cause which they are supposed to pursue. This is one of them.


Ontario’s anti-carbon-tax-sticker legislation, enacted in a perversion of parliamentary democracy, is likely unconstitutional, and wrong in principle. The day when it is repealed on struck down by the courts cannot come too soon. It might seem like a small thing― it’s just stickers at gas stations, after all, and unlike with the various recent “statements of principles” and “attestations” nobody is required to believe, or even pretend to believe, what the government wants them to say. Nevertheless the impulse behind this legislation is not that much less authoritarian than that behind these other denials of the freedom of speech.

This is a reminder that liberty is under threat both from self-styled progressives and from self-anointed populists. Each camp will happily point to the other’s excesses and may even proclaim itself a defender of rights, freedoms, and the Charter; both are hypocrites. It is essential that firm, neutral principles of freedom be upheld against threats on either side.

Modern Mailmen

Back-of-the-envelope thoughts on what the history of postal services and their competitors can teach us about the regulation of social media

This post is co-written with Akshaya Kamalnath[*]

One of us (Akshaya) recently visited the Postal Museum in Washington DC. Looking at the historical development and role of the postal services in the US brought to mind our modern forms of communication—social media platforms—and their value, especially in terms of free speech. We often associate free speech with the press but, as a quote by Nat Hentoff in the Postal Museum informs visitors, it was the post that brought news to the press and then brought newspapers to the public.

Today, social media platforms like Facebook, Twitter and Google (YouTube) play a role similar to that of the postal service by acting as intermediaries for communication. They are, in a sense, the high-tech descendants of the postal services. The post physically transports letters and parcels from one person to another, while Facebook electronically transmits speech that one person wants to convey to others. The tech platforms have just made it easier to convey messages to a number of people at once. Tech platforms also help transmit news content—just like the post delivers newspapers. In fact, the use of postal services to deliver newspapers was considered the most important information technology in the late 1700s.  

As lawmakers are talking about regulating speech on social media platforms, a comparison with postal services is instructive. The postal service is not required or even allowed to scrutinize people’s mail and make decisions about whether or not to deliver it. So why should its technologically more advanced relatives have to identify and remove misinformation or statements supposed to be “hate speech”? Of course, social media can be used to commit crime, including engaging in hate speech as defined in the criminal law of some countries including Canada. The post collaborated with law enforcement where necessary to investigate fraud and other criminal activities and social media companies should do the same. Social media companies should obviously comply with court orders if someone is found to have committed a crime. The issue is whether they should be expected to engage in preventive enforcement.

The further question about whether we should require these tech platforms to service all users equally, like the postal service is expected to, is more complicated. This is because the dominant postal service is usually run by the state, while the tech platforms like Facebook are run by corporations in the private sector. While we can ask a state-run enterprise to provide services to all equally, more thought needs to be given before private enterprises are held to the same standard. Yet, government regulation is being considered because, among other things, there are complaints about the spread of what activists deem to be “hate speech”, and also complaints about the silencing of conservative voices on social media.

Overall, we have to tread carefully with government regulation. In addition to interfering with their freedom of expression and association, heavy-handed regulation of online platforms would have the effect of making it harder for new and, at least initially, smaller players to enter the social media market, which ought to be the real solution to the concerns about the existing platforms’ behaviour. It will be highly unlikely that we see university students create the next Facebook or Google from their dorm rooms or garage if regulation becomes burdensome.

The history of the postal services can again serve as a warning to resist government-backed monopolies, which Facebook and the few other social media giants can in effect become if government regulation becomes burdensome. It is telling that the Postal Museum makes no mention of Lysander Spooner who tried to set up a private postal service in 1844.

Spooner argued against state monopoly over the postal service, saying:

The present expensive, dilatory and exclusive system of mails, is a great national nuisance—commercially, morally and socially. Its immense patronage and power, used, as they always will be, corruptly, make it also a very great political evil.

He added (referring to the US Constitution’s First Amendment protection for free expression) that

any law, which compels a man to pay a certain sum of money to the government, for the privilege of speaking to a distant individual, or which debars him of the right of employing such a messenger as he prefers to entrust with his communications, “abridges” his “his freedom of speech”.

Although Spooner’s business was eventually forced to close by a tightening of legislative protections for the government post’s monopoly, it had the temporary impact of bringing down the cost of postal services.

Government regulation requiring Facebook and other social media platforms to set up a complex decision-making system to enforce restrictions on what messages they can be used to convey will increase the cost of operating such platforms. Any new platform will be required to spend heavily on human moderators, artificial intelligence systems capable of assisting them, or, likely, a combination of the two. While established platforms like Facebook will not find it difficult to invest in complying with such regulations, the cost will be prohibitive to outsiders who want to set up competing social media platforms. This should explain why Mark Zuckerberg, CEO of Facebook, is in favour of government regulation.

Heavy regulation of speech on social media also runs the risk of government using social media to their political advantage—a modern version of the political abuses of the power over the transmission of ideas that Spooner denounced, which we are already seeing in some countries. In France, it emerged that the President’s office circulated a doctored video on social media, despite the President himself being committed to censorship of “fake news”. In Austria, a politician asked Facebook to take down a post calling her a “lousy traitor of the people”, a “corrupt oaf” and a member of a “fascist party” none of which amounted to hate speech under Austrian law.

The converse possibility, regulation requiring Facebook and other platforms to host all users irrespective of their opinions, would also be problematic, because it would infringe the platforms’ ability to hold and act on their own views, as well as to provide an environment in which they think their customers will be happiest. Just like restaurants may ask misbehaving patrons to leave so others may enjoy their dinner, social media platforms should be able to decide where to draw the line so that a large majority of their users are able to enjoy the platform. Or, to return to the postal analogy, suppose a private delivery company insisted on reading the letters or examining the content of the packages we wanted it to deliver for us and refuse to deliver those it deemed morally objectionable. The appropriate response for a person who did not want his or her letters read, or who submitted to the exercise and had a letter rejected, would be to go to a competitor—or to establish one, like Spooner did—rather than to force his message on a party unwilling to deliver it. Similarly, when Facebook or other online platforms set out standards with regard to the type of content and members it will allow, they make specific choices as private actors, and should be free from the government’s interference.

All this is not to say that the large social media platforms should do nothing to address the problems associated with their use. Companies like Facebook are under pressure from their shareholders and consumers. Facebook’s shareholders recently demanded a change in management since the current management had not dealt with misinformation and hate speech. Even though Mark Zuckerberg holds the majority voting power in the company, the shareholder proposals convey a message. Facebook’s management is aware of the market pressures and has taken a number of measures, including releasing its public content-moderation rules and a proposal for an independent body to hear appeals regarding decisions by Facebook regarding content moderation. (That said, presumably, the independent body would still be working under guidelines that Facebook has drafted or at least is in agreement with.)

While not perfect, these are voluntary responses to market sentiment against problems of misinformation and censorship that big social media companies have chosen to invest in. Facebook’s taking such measures does not preclude a new company from starting a modest platform without having to invest in these systems at the outset. As they get bigger, the new competitors could devise their own solutions on different principles, rather than having to follow a pattern imposed by legislation, not only enacted at Facebook’s suggestion but, quite possibly, drafted based on its proposals. Just like new courier companies have differentiated themselves from postal services based on GPS tracking, expedited delivery or convenient package pick-up options, new social media companies may exploit gaps especially if the big social media companies preclude certain views on their platforms.


[*] Dr Akashaya Kamalnath is a corporate and insolvency law scholar. She is currently teaching at Deakin University, but will be joining the Auckland University of Technology Law School shortly. You can read her papers here, and follow her on Twitter.