Offspring of Depravity

The origins of the administrative state, and why they matter

To a degree that is, I think, unusual among other areas of the law, administrative law in the United States and, to a lesser extent, in Canada is riven by a conflict about its underlying institution. To be sure there, there are some constitutional lawyers who speak of getting rid of judicial review of legislation and so transferring the constitution to the realm of politics, rather than law, but that’s very much a minority view. Labour unions have their critics, but not so much among labour lawyers. But the administrative state is under attack from within the field of administrative law. It has, of course, its resolute defenders too, some of them going so far as to argue that the administrative state has somehow become a constitutional requirement.

In an interesting article on “The Depravity of the 1930s and the Modern Administrative State” recently published in the Notre Dame Law Review, Steven G. Calabresi and Gary Lawson challenge the defenders of the administrative state by pointing out its intellectual origins in what they persuasively argue was

a time, worldwide and in the United States, of truly awful ideas about government, about humanity, and about the fundamental unit of moral worth—ideas which, even in relatively benign forms, have institutional consequences that … should be fiercely resisted. (828)

That time was the 1930s.


Professors Calabresi and Lawson point out that the creation of the administrative state was spearheaded by thinkers ― first the original “progressives” and then New Dealers ― who “fundamentally did not believe that all men are created equal and should democratically govern themselves through representative institutions”. (829) At an extreme, this rejection of the belief in equality led them to embrace eugenics, whose popularity in the United States peaked in the 1930s. But the faith in expertise and “the modern descendants of Platonic philosopher kings, distinguished by their academic pedigrees rather than the metals in their souls” (829) is a less radical manifestation of the same tendency.

The experts, real or supposed ― some of whom “might well be bona fide experts [while] [o]thers might be partisan hacks, incompetent, entirely lacking in judgment beyond their narrow sphere of learning, or some combination thereof” (830n) ― would not “serve as wise counselors to autonomous individuals and elected representatives [but] as guardians for servile wards”. (830) According to the “advanced” thinkers of the 1930s, “[o]rdinary people simply could not handle the complexities of modern life, so they needed to be managed by their betters. All for the greater good, of course.” (834) Individual agency was, in any case, discounted: “the basic unit of value was a collective: the nation, the race, or the tribe. Individuals were simply cells in an organic whole rather than ends in themselves.” (834)

Professors Calabresi and Lawson are careful to stress that the point of their argument is not condemn the administrative state by association with the worst excesses of the times in which it originated. Rather, they want to push back against the trend, exemplified in articles such as Gillian Metzger’s “1930s Redux: The Administrative State Under Siege“, of treating the foundation of the administrative state as deserving of particular deference or respect. They explain that

[b]ecause there is no authoritative constitutional text emanating from the 1930s, any reasons for treating that decade as interpretatively sacrosanct must focus on the moral goodness of the ideas that grounded that period. Many of the intellectual currents that dominated the 1930s were, frankly, very bad. As a starting point for thinking about human affairs, one’s first instinct should be to run as far away from that decade as quickly as one can. More fundamentally, the bad ideas of the 1930s that specifically drove the construction of certain parts of the modern administrative state—belief in omnipotent government by socially superior experts under broad subdelegations of legislative power, with a formal (or rote) separation of powers seen as an anachronistic hindrance to modern scientific management of people, who are not ends in themselves but simply means to the accomplishment of collective nationalist or tribalist ends—are at the intellectual core of just about everything bad that occurred during that decade. (839)

Professors Calabresi and Lawson conclude that, instead of looking to the 1930s as a source of public law we should ― even on purely moral grounds, in addition to fidelity to law ― we should look to the 1780s and the 1860s. The former decade was marked by “libertarian and egalitarian commitments to replace European feudalism with something new and better”, (842) as well as to separation of powers; the latter, by important progress in the implementation of those libertarian and egalitarian commitments, initially admittedly honoured in the breach in many ways. Professors Calabresi and Lawson also appeal to another historical point: the signing of the Magna Carta at Runnymede in 1215, to which they trace what they call “the principle of legality, which says that executive and judicial actors can only act in accordance with preexisting law”. (863)


While I think it is a little, and perhaps more than a little, optimistic to connect this principle ― this formulation of the Rule of Law ― to the Magna Carta, it is supposed to be central to Canadian, and not only American, administrative law. As the Supreme Court said in Dunsmuir v New Brunswick, 2008 SCC 9, [2008] 1 SCR 190, “[b]y virtue of the rule of law principle, all exercises of public authority must find their source in law.  All decision-making powers have legal limits”. [28] But the belief in the superiority of administrative power wielded by alleged experts for what is deemed, by them, to be the public good is very much a part of our administrative law too, and it goes back to the same roots as that of the American champions of the administrative state. As co-blogger Mark Mancini has argued here,

the reasons marshalled for why we defer to administrative agencies are the same today as they were in the 1940s. … For the most part, Canadian administrative law continues to be stuck in the thrall of American Progressivism—by which I mean [the] school of thought[] dominant in the New Deal era.

As Mark notes, “in Canada, we had our own band of administrative law Progressives” ― though of course they looked to the United States for inspiration. (There’s anything wrong with looking to the United States, of course; that’s what I’m doing here!) But then again, we had also had our own band of eugenicist progressives too, some of whom have statues on Parliament Hill. And we had our more peculiar rotten ideas about government too. The 1930s were a bad time ― arguably an especially bad time― in Canada, as well as in the United States and, for this reason, the argument made by Professors Calabresi and Lawson is relevant to Canadians.

Of course, the Canadian constitution is not the same is the American one. In particular, it does not incorporate as strong a conception of the separation of powers. Arguments to the effect that the administrative state in its current form is unconstitutional are much less straightforward in Canada; perhaps they are wrong. Certainly the case against the delegation of legislative power is more difficult to make under the Constitution Act, 1867, than under the U.S. Constitution. But all this means is that the moral case made by Professors Calabresi and Lawson is that much more significant. If the modern administrative state is the misbegotten offspring of an especially depraved epoch, then it should be dismantled, even if it is not unconstitutional. (The case for it being constitutionally required, however, is that much weaker ― not that it had much strength to begin with.)

And the advice to look to the 1780s or the 1860s is applicable to Canada too. Admittedly, the 1780s do not hold the same significance for our constitutional history as they do for our neighbours. But the ideas of what Jeremy Waldron calls “enlightenment constitutionalism”, which Professors Calabresi and Lawson associate with the 1780s, are relevant to Canada. Indeed, our own constitutional arrangements implement some of what, as I suggested in my critique of Professor Waldron’s arguments here, were the Enlightenment’s signal contributions to constitutional thought ― federalism and judicial review of legislation. As for the 1860s, sapienti sat.


As I noted at the outset, the moral worth of the administrative state is not just a matter for political philosophers to debate. It is an issue that is tied up with the ongoing fights about the details of administrative law doctrine. Perhaps this worth is unconnected to its sinister origins. But I think that it is for pro-administrativists to make this case. And I am quite skeptical that they can succeed. As have noted a number of times, most recently here, “[t]he administrative state is the state of prisons, of border control, of professional regulators determined to silence their members if not to impose official ideology on them”. It has come rather less far from its smug, authoritarian beginnings than its defenders would have us believe.

The Five-Judge Myth

How many Supreme Court judges does it take to decide a civil law appeal?

By Peter McCormick

A defining aspect of the Canadian legal system is its bijuralism: Quebec’s civil law system is distinctly different from the English-derived common law of the other provinces. The federal-provincial division of powers which assigns to the provincial legislatures jurisdiction over “property and civil rights within the province” is its formal entrenchment. There are also structural accommodations, one of the most important of which is the composition of the Supreme Court.  Alone among the provinces, Quebec is guaranteed a minimum share of the Court’s membership. One third of the judges (two of six in 1867, three of nine since 1949) must be appointed from the bar or the judiciary of Quebec, which is to say that they must be experienced in the civil law.

On the face of it, this is not enough; it does not preclude the possibility of a common law majority that persistently out-votes its civil law minority and steadily erodes this bijuralism. Its impact has therefore been reinforced by a long-established practice. Peter Hogg describes it as follows: “since 1949 … it has been possible to assemble a quorum of five judges with a majority of civilians” with the result that “(t)his is now the usual composition of the bench when the Court hears a civil law appeal from Quebec.” (Hogg, Constitutional Law of Canada (various editions), Chapter 8.5(a).) Assuming a unified trio of civilian judges – a single defection can be decisive – it is both an acknowledgement and an effective protection of Quebec’s civil law uniqueness.  The recent controversy over the Nadon appointment served once again to highlight the importance of demonstrable and recent civil law experience for those Quebec judges.  As a student, years ago, I was impressed by the elegance of this “five-judge” solution; as a professor, I tried to ensure that my students appreciated it as well.

However, there is another story that we have been telling about the Supreme Court, and that is the story of a steady move toward larger panels.  The Supreme Court Act permits panels of various sizes, but five judge panels continued to dominate even after the enlargement of the Court to nine members in 1949.  In this context, a slight tweak of the rules for striking the panels for civilian appeals was procedurally simple, almost invisible, and reliably consequential.  Ever since the great watershed of the Laskin Court, however, panels have been getting steadily larger.  On the Lamer Court, the default was already seven judges, with the more important issues (such as the growing number of constitutional cases) assigned to larger panels and only the more routine cases (such as appeals by right) going to smaller ones.  Under McLachlin, this trend has continued, such that nine-judge panels are now the most common and five-judge panels have become unusual, used for only one reserved judgement in every thirty.


It is not easy to reconcile this long-term trend toward large panels with a five-judge rule for civil appeals.  This post reports on my own investigation of these two on-the-face-of-it contradictory generalizations, focusing initially on the McLachlin Court.  On my findings, it is the “larger panels” generalization that very much prevails.  The “five-judge” practice of Quebec exceptionalism has all but disappeared.

The first question is how to objectively identify the set of civil law appeals, and the Supreme Court itself has provided the most obvious solution: the judge-written headnotes that lead off every decision.  If those included specific mention of either or both of the Civil Code of Quebec and the Code of Civil Procedure, then I treated it as a civilian appeal.  (A further forty cases listed these statutes among their citations without any headnote notation, but I did not treat citation alone as justifying their inclusion.)  Limiting the inquiry to reserved judgments only, this gave me fifty-five civil law cases, for an average of about three per year.

How many of these were decided by five judge panels?  Only five – one in every eleven, which is to say one every three or four years.  Thirty-two went to seven judge panels, and nineteen to full-court nine judge panels. The average panel size was 7.5, only slightly below the McLachlin Court average of 7.9 for all reserved judgments.  The fact that it is lower at all may suggest a residual tug of the older “five-judge” rule, but if so it is a small tug indeed.

Even more surprising, only a single one of those five judge panels included all three Quebec judges, guaranteeing that a united set of Quebec civilians would prevail over their common law colleagues.  More remarkably yet, this was balanced by a single example at the opposite extreme — a panel with no Quebec judges at all.  A panel small enough that the Quebec judges can make up a majority is of course also small enough that the Quebec judges can be left out altogether.  The five-judge rule would have led us to expect that these five panels would have included a total of fifteen Quebec judges and ten of their common law colleagues; in practice, they included only eight, well below the common law total of seventeen.  Further to punctuate the point, four of the five examples were from the first four years of the McLachlin Court, and the single more recent example was the “no Quebec judges” panel.


Comparing eighteen years of McLachlin with eighteen months of Wagner calls for caution, but there has been no sign of a reversal of the above patterns.  To date, the Wagner Court has dealt with seven civilian appeals, some of which were consequential; five were decided by panels of nine and two by panels of seven.  There was no sign of the five-judge practice, no indication that these appeals are treated differently in this respect from the broad run of reserved decisions.  The five-judge rule is dead; it seems to have breathed its last in 2004.

But all is not lost.  Quebec judges may have been under-represented on the vanishing smaller panels and risk being outvoted on the larger ones, but they do deliver most of the judgments – fifty-one of the McLachlin Court’s fifty-five and five of the Wagner Court’s seven for an overall total of fifty-six out of sixty-two, about ninety per cent. There has long been a significant “homer” tendency on the Supreme Court in assigning the judgment – an appeal coming from your own province roughly doubles your chances – but the tendency is even stronger for Quebec civil appeals.  Compared with the five-judge rule, this may well be a less robust and less compelling institutional recognition of Quebec exceptionalism, but it is where the empirical evidence takes us.  We should remember, however, that when the Supreme Court was first established it was the spectre of common law judges deciding civil code issues that worried Quebec. “It only happens one time in every ten” may not be a completely reassuring response now that civilian judges are now outnumbered on every panel.

Shouting into the Constitutional Void

Section 28 of the Canadian Charter and Québec’s Bill 21

By Kerri A. Froc*

“And if thou gaze long into an abyss, the abyss will also gaze into thee.” (Friedrich Nietzsche, Beyond Good and Evil. Aphorism 146)

For several years now, I have been arguing that section 28 of the Canadian Charter of Rights and Freedoms is more than a symbolic flourish, more than just emphasis for section 15’s sex equality guarantee, and more than an interpretive provision.  In fact, it has its own independent work to do.  This includes blocking attempts by government to use section 33 to preserve gender inequality. 

I did not make up this interpretation of section 28.  Rather, it is part of section 28’s text and history and is uncontroversial amongst those who have studied the matter.  That is why I am not only perplexed, but annoyed, at section 28 seemingly being ignored in the debate over the constitutionality of Bill 21’s requirement that certain government employees (including school teachers, police, Crown prosecutors and judges) do not wear religious symbols at work (section 6).  It is in fact reminiscent of the way that women’s rights were ignored in 1981 constitutional negotiations, which galvanized women to insist upon section 28 in the first place.  Below, I discuss section 28’s interpretation vis a vis section 33, and then how it would be pled in a constitutional challenge to Bill 21.


Section 28 beginning phrase reads: “Notwithstanding anything in this Charter.”  This meant its guarantee of equal rights is not to be derogated by other provisions of the Charter. Provincial and federal bureaucrats attempted after the November 1981 “Kitchen Accord” to subject section 28 to section 33.  They drafted amendments to section 28 and section 33, notionally to “implement” the terms of the Accord (though first ministers never discussed section 28).  The opening words of Section 28 would have been revised to read, “Notwithstanding anything in this Charter except section 33,” and section 33 would have been amended to end with, “or section 28 of this Charter in its application to discrimination based on sex referred to in section 15.”  These proposed additions were scrubbed from the Charter’s final text through the hard work of feminist advocates, women MPs from all parties, and, to put it bluntly, a groundswell of pissed off women from across the country.  This history, however, merely confirms that “notwithstanding anything” means what it plainly says.

In their 1984 book, Canada Notwithstanding, Roy Romanow, John Whyte and Howard Leeson (all members of the November 1981 Saskatchewan constitutional delegation) confirmed that the removal of the application of section 33 from section 28 “in effect…meant that sexual equality in section 15 could not be overridden.”  Justice Carole Julien, in a 2004 Charter case involving pay equity, Syndicat de la fonction publique c. Procureur général du Québec,had occasion to discuss the legal effect of section 28.  She noted that the predominant scholarly opinion was that the override did not apply to section 28 “due to the historical context of its adoption and its objectives” (my translation).  It is unfortunate that this judgment was merely a passing footnote in the recent Supreme Court decision, Centrale des syndicats du Québec v. Quebec (Attorney General).


How would it potentially play out if litigants argued section 28 in relation to the Bill 21 constitutional challenge?  There are potentially two Charter claims that could be advanced by women who are adversely affected by section 6.  The first is that it discriminates against them on the basis of sex, contrary to section 15(1).  The second is that section 6 violates their freedom of religion disproportionately, so that women are unable to exercise this freedom on an equal basis with men.  Sex discrimination is contrary to Charter section 15(1) and 28; a gender-disproportionate violation of religious freedom would be contrary to sections 2(a) and 28.  Section 28 is involved in both claims as section 6 results in unequal rights afforded to men and women.   A section 28 violation cannot be preserved using section 33.

One could also use an alternative legal argument in relation to section 15.  Quebec could argue that a general sex equality violation, in and of itself, does not implicate section 28 (saying that section 28 does not really “add” anything to the section 15 determination).  However, if additional state action is taken to attempt to preserve a section 15 sex equality violation by invoking section 33, section 28 operates to block the effect of that invocation.  Taking action to preserve women’s section 15 rights violation results in unequal rights contrary to section 28.  This is quite applicable to Bill 21, in that section 30 contains a pre-emptive declaration that the Act operates notwithstanding sections 2 and 7-15 of the Charter. 

Regardless of which argument(s) you accept, the validity of section 6 cannot be maintained by the section 33 override because doing would mean section 28 is made subject to the legal effect of section 33.


A question I am sometimes asked is: where is the gender inequality in Bill 21?  Many media sources have indicates that the group most affected are Muslim women wearing the head scarf (hijab), but do not indicate the sources they rely upon for that fact.  I’ve done some of my own data crunching to provide initial support for that point. 

Of the groups mentioned, Muslims are in vastly greater numbers in Québec than both Jews and Sikhs (men from these two other groups have been mentioned as being the others affected by the law).  For the last year in which we have data (2011), there were nearly two and a half times as many Muslims in Quebec as Jews and Sikhs together. Approximately 53%, of Muslim women in Canada wear the hijab.  Quebec’s public service is still massively dominated by white francophones; however, nearly half of its workers are female (amongst school teachers, one of the largest groups affected by Bill 21, that percentage is much higher). It stands to reason given these statistics that most of those affected are Muslim women.  While some judges may not consider these statistics more than a “web of instinct”, this data could be supplemented by access to information requests and litigation disclosure to obtain numbers of affected employees.  Further, one could argue that the state demanding women remove clothing has a more threatening import and communicates a sex-specific devaluation, given the way women’ attire has been regulated and judged by law throughout history.  Thus one could argue that the qualitative impact constitutes a sex-based distinction in itself. 

Even apart from disparate impact, if the purpose of a law is discriminatory or is to privilege certain religious beliefs, then that would be a violation of section 15(1) and section 2(a) respectively.  A good case could be made that Bill 21 targets Muslim women based, for instance, on the Quebec Minister for the Status of Women’s comments.  Concerning the privileging of religious beliefs, it is worth noting that symbols of Quebec’s “religious cultural heritage” (read: Christianity/Catholicism) are specifically exempted from all of Bill 21’s provisions by section 16. 

Of course, there are potentially other elements in relation to a Charter analysis that would have to be successfully argued, such as showing “disadvantage” for section 15(1) and more than atrivial infringement of religious freedom, for section 2(a).  However, I do not regard those as posing much of an impediment. 


Why should we care if civil liberties associations, lawyers, and courts ignore section 28 in the upcoming constitutional battle over Bill 21?  To paraphrase Nietzsche, if we gaze into the Constitution and see only an abyss when it comes to section 28, we should not be surprised if the abyss gazes back in the form of more constitutional provisions courts feel secure in being able to ignore into desuetude.  Simply put, entrenched constitutional text should and does count more than implied bills of rights, unwritten principles, constitutional architecture and the like.  If not section 28 in this case, then when?


* Kerri A. Froc is an Assistant Professor in the Faculty of Law, University of New Brunswick. Follow her on Twitter!

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.

It’s Happening Here Too

Canadians need to heed David Bernstein’s warning about administrative decision-makers’ disregard of constitutional rights

A very interesting article by David E. Bernstein, “Anti-Discrimination Laws and the Administrative State: A Skeptic’s Look at Administrative Constitutionalism” has recently been published in the Notre Dame Law Review. Professor Bernstein cautions against allowing administrative decision-makers to pursue egalitarian goals unchecked by judicial supervision, because this pursuit often tramples over constitutional guarantees, especially freedom of speech. It is a compelling warning, and deserves the interest of Canadian readers, because the problems Professor Bernstein identifies afflict Canadian law. Indeed, much of his argument applies to the administrative enforcement of other statutes, not only anti-discrimination ones.


Professor Bernstein takes aim at the view, which he attributes to a significant number of American scholars, that administrative decision-makers both do and ought to play a very significant role in defining the scope and content of constitutional protections for certain fundamental rights. This view, “administrative constitutionalism”, rests on a number of arguments. Its supporters think that administrative decision-making “is more transparent than” the judicial sort, that administrators “are more accountable to public opinion than are courts”, and that they bring their expertise to bear on the application of constitutional standards to particular regulatory schemes. (1384) Professor Bernstein provides a number of examples of administrative decision-makers “aggressively enforcing antidiscrimination laws at the expense of constitutional protections for freedom of expression and guarantees of due process of law”, (1386) sometimes in defiance of relevant Supreme Court precedent and political direction. These will be of considerable interest to readers who follow American legal and political developments.

But what is more interesting from a parochial Canadian perspective is Professor Bernstein’s analysis of the situation ― his explanation for why administrative decision-makers tend to apply the law in a way furthers their statutory mission at the expense of the constitutional rights of those subject to their decisions. The explanation is partly institutional, and partly ideological.

The first institutional fact that contributes to administrative disregard of constitutional rights, according to Professor Bernstein, is that administrative decision-makers “maximize their power and budget”, and secure “political support, by expanding the scope of the laws they enforce”. (1401) Constitutional limits to this expansion are brushed aside. Second, a purposivist approach to statutory interpretation “practically invites agencies to find and even create ambiguities so that they can interpret statutes broadly”. (1402) In doing so, administrative decision-makers see themselves as accomplishing legislative goals, and ignore the compromises that may have been involved in the enactment of their enabling legislation. Third, administrative “agencies tend to attract employees who are committed to the agency’s regulatory mission” (1403) and want to expand their own power to, as they see it, do good. While some instances of regulatory overreach invite pushback from those subject to the regulation, this is generally not the case when it comes to “antidiscrimination regulation”, in part because “many businesses hesitate to publicly oppose” this regulation “because of the negative public relations implications”. (1403) Fourth and last, administrative decision-makers “do not see enforcing constitutional constraints on their authority as their job”. (1404) The courts themselves are partly to blame for this, because they often discourage the bureaucrats from looking to the constitution. But, for their part, supporters of “administrative constitutionalism” positively encourage administrative decision-makers to treat constitutional constraints as no more than a factor, among others, to take into account or to reject.

As for ideological concerns, they have to do with the fact that “conflicts between freedom of expression on the one hand, and restrictions on discrimination by private actors on the other, are conflicts between a
constitutional right and a statutory privilege”. (1406) As a matter of orthodox law the former ought to prevail, but for those “who believe that protecting vulnerable groups from discrimination should be at the heart of our legal and political system”, (1406) such an outcome would be wrong. They are accordingly inclined to discount constitutional concerns, or to seek to re-balance them by appealing to “the notion that the ‘constitutional value’ of antidiscrimination should trump First Amendment limitations on government regulation”. (1407) These views are prevalent not only in the legal academy, but also among activists ― and their ideological allies among the administrative decision-makers in charge of enforcing anti-discrimination laws. The fact that “[a]t the state and local level” these decision-makers are often

known as ‘human rights commissions’ … suggest[s] that the right to be free from private discrimination is at least as valuable as other rights, including constitutional rights. Indeed, the phrase ‘human rights’ suggests a superiority over mere textually supported constitutional rights. (1408)

So why, Professor Bernstein asks, don’t the courts do something about administrative decision-makers run amok? After all, the courts ― at least “generalist courts” ― “do not share mission-driven agencies’ tunnel vision, i.e., the latter’s devotion to its statutory mission at the expense of
other considerations”. (1410) But the administrative state is often able to escape scrutiny by using settlements or ostensibly “soft” forms of regulation that are not subject to judicial review. Professor Bernstein argues that courts should engage in review of administrative action more often, and that they ought to be less deferential when they do so. He also suggests possible institutional reforms, notably “to establish constitutional watchdog offices devoted to protecting constitutional rights from
[administrative] overreach”, (1413) whether within individual administrative entities or for the government as a whole.


Canadian readers probably do not need me to tell them that the issues Professor Bernstein describes arise with at least as much, and probably more, urgency in Canada. After all, although it rests on foundations that are partly different from those of its American counterpart, and goes by a different name, administrative constitutionalism is the law of the land in Canada, whenever a court is minded to follow the precedent set in Doré v Barreau du Québec, 2012 SCC 12, [2012] 1 SCR 395. In Doré, the Supreme Court held that, given their alleged expertise in applying constitutional “values” in the context of specific statutory schemes, administrative decision-makers are entitled to judicial deference, even in cases where the Canadian Charter of Rights and Freedoms is implicated. Whether an administrative decision gives effect to constitutional “values” ― not even rights ― as fully as possible in light of the statutory objectives is to be assessed on a standard of reasonableness. The Supreme Court also confirmed that reasonableness is the presumptive standard of review applicable to the decisions of anti-discrimination tribunals, in Canada (Canadian Human Rights Commission) v. Canada (Attorney General), 2018 SCC 31, [2018] 2 SCR 230 (although this was not a Charter case).

Admittedly, the Supreme Court hasn’t always been inclined to do so, occasionally simply ignoring Doré. But its latest engagement with administrative interference with constitutional rights, in Law Society of British Columbia v Trinity Western University, 2018 SCC 32, [2018] 2 SCR 293, reiterated the applicability of the Doré framework, although it is worth pointing out that the Court’s majority insisted that this wasn’t supposed to be “a weak or watered-down version of proportionality”. [80] Still, the majority wrote that

Doré’s approach recognizes that an administrative decision-maker, exercising a discretionary power under his or her home statute, typically brings expertise to the balancing of a Charter protection with the statutory objectives at stake … Consequently, the decision-maker is generally in the best position to weigh the Charter protections with his or her statutory mandate in light of the specific facts of the case … It follows that deference is warranted when a reviewing court is determining whether the decision reflects a proportionate balance. [79; references omitted]

Professor Bernstein’s article helps us identify the folly of this approach. Despite the claims to the contrary of Justice Abella (the author of the Supreme Court’s opinion in Doré and the most strident defender of “administrative constitutionalism”, most recently in Canada (Public Safety and Emergency Preparedness) v Chhina, 2019 SCC 29) and her colleagues, administrative decision-makers are unlikely to take the constitution, or even constitutional “values”, seriously at all. Granted, unlike their American counterparts, Canadian courts do not discourage bureaucrats from taking the Charter into account. Justice Abella, in particular, exhorts them to do so. But such exhortation is unlikely to mean much, compared with the much more concrete incentives Professor Bernstein identifies.

Canadian bureaucrats, no less than their American colleagues, want to expand their power and to advance their and their allies’ ideological goals. The seemingly expanding efforts of human rights bureaucracies or other administrative decision-makers (such as the former benchers of the former Law Society of Upper Canada) to police speech in the name of equality are an illustration of these twin tendencies. And while there has been pushback against the Law Society’s demand that lawyers “promote equality, diversity, and inclusion”, culminating in the election of a plurality of benchers opposed to this imposition, the incentives, both in the private sector and, still more in, say, public educational institutions are very much on the side of tacit or even vocal endorsement of the one-way ratchet of obstensibly pro-equality agenda.

The Supreme Court’s rulings on statutory interpretation exacerbate this problem. In West Fraser Mills Ltd v British Columbia (Workers’ Compensation Appeal Tribunal), 2018 SCC 22, [2018] 1 SCR 635, the majority insisted that the statute at issue featured a “broad and unrestricted delegation[s] of power” [11] so that an administrative decision-maker could pursue its purposes; in TWU, the majority also spoke of a statutory objective “stated in the broadest possible terms”. [33] (West Fraser, to be sure, was not a case implicating constitutional rights. TWU was such a case, however, and their logic is much the same.) In both cases, as I explained respectively here and here, the majority gave no effect to statutory language suggesting that the administrative decision-makers’ powers were not, in fact, unlimited, to which dissents sought to draw its attention. In West Fraser, the majority opinion disparaged attention to such details as “formalistic”. [18] As Professor Bernstein points out, when empowered to pursue expansively defined statutory missions, administrative decision-makers will be unlikely to pay much heed to constitutional concerns. Indeed, TWU offers a perfect illustration of this, since the Supreme Court ended up having to make up the reasons that supposedly justified the administrative decisions at issue.

What Professor Bernstein terms “ideological” factors operate in Canada too. Here too, the value of non-discrimination in the private sphere, branded as a “human right” by federal and provincial legislation alike is held to prevail over such constitutional concerns as freedom of expression and freedom of conscience. The TWU majority spoke of “shared values”, notably non-discrimination, as a valid reason for limiting constitutional rights, despite the fact that the Charter explicitly provides that it does not expand the law-making powers of legislatures or their creatures in the administrative state.


Like Professor Bernstein, I will conclude with an appeal for greater judicial scrutiny of administrative decisions that implicate constitutional rights. Judges ought to realize that administrative decision-makers have no particular incentive to be mindful of the constitution, and real incentives to disregard it. Even when they act in good faith, bureaucrats suffer from a single-minded, excessive focus on their statutory missions, real or assumed, that is bound to divert their attention from constitutional rules that ought to be paramount for all those who exercise public power, but in reality matter primarily to the courts ― if they matter to anyone.

To be clear, the issue is not only with the Doré framework ― though this is the most obvious way in which excessive and unwarranted deference is given to administrative decision-makers when they decide Charter questions. The Doré framework must go, the sooner the better, but this is not enough. The idea that “values” are an adequate substitute for law, whether as a source of constitutional guarantees or of administrative powers, must go along with the Doré framework, to which it is closely linked. And the Supreme Court’s approach to statutory interpretation, and in particular its willingness to countenance supposedly “unrestricted delegations” of power to administrative decision-makers, even if this requires disregarding more circumscribed statutory language, must go too. This, in turn, may require an end of the Court’s fascination with administrative expertise and its pro-regulatory bias.

This is, admittedly, a very ambitious programme. But, as Professor Bernstein shows, it is on that must be attempted if constitutional constraints are to be meaningful in the administrative state. “Administrative constitutionalism” is no substitute for the real thing. This is precisely why its supporters, who are not willing to accept constraints on what they believe is the bureaucracy’s power to do good, advocate for it. This is why we must reject it.

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.