The Rule of Law All the Way Up

Introducing my recently-published chapter on the Rule of Law and Canadian constitutional law

LexisNexis Canada recently published (if I understand correctly, as a standalone book as well as a dedicated issue of the Supreme Court Law Review (2d)) Attacks on the Rule of Law from Within, a collection of essays co-edited by my friends Joanna Baron and Maxime St-Hilaire. The publisher’s blurb gives a concise summary of the project’s background and contents:

This volume is a collection of six papers developed from the Runnymede Society’s 2018 national conference by a community of legal experts in response to Supreme Court of Canada Justice Rosalie Abella’s comment that “the phrase ‘rule of law’ annoys her”. 

Grounded on the intuition that the legal profession supports the rule of law, the papers examine the historical perspective on threats to the rule of law, the sufficiency of the current Canadian legal framework to support this ideal and how the principle of stare decisis as observed by the Supreme Court of Canada undermines the spirit of the rule of law. The volume also discusses how the law relating to Aboriginal title and the duty to consult fails to adhere to the Rule of Law standards … to the detriment of indigenous and non-indigenous Canadians alike.

I am honoured to have contributed to this volume, with an essay called “The Rule of Law All the Way Up”, which focuses on what I see as the lack of commitment to the Rule of constitutional Law in by scholars, judges, and politicians. Here is the abstract:

Canadian constitutional law is seldom criticised for its failure to live up to the ideal of the Rule of Law. This article argues that it should be so criticised. A number of widely accepted or uncontroversial Rule of Law requirements―the need for general, stable, and prospective rules, the congruence between the “in the books” and the law “in action, and the availability of impartial, independent courts to adjudicate legal disputes―are compromised by a number of ideas already accepted or increasingly advocated by Canadian lawyers, judges, and officials.

This article describes four of these ideas, to which it refers as “politicization techniques”, because they transform what purports to be “the supreme law of Canada” into a set of malleable political commitments. These are, first, deference to legislatures or the application of a “margin of appreciation” and the “presumption of constitutionality” in constitutional adjudication; second, constitutional “dialogue” in which courts not merely defer, but actively give way to legislative decisions; the substitution of political for legal judgment through the application of the “notwithstanding clause” of the Canadian Charter of Rights and Freedoms; and the rewriting of constitutional law by the courts under the banner of “living tree” constitutional interpretation.

The article concludes with an appeal to those who profess commitment to the Rule of Law in relation to the Constitution not to embrace or endorse the means by which it is subverted.

The entire chapter is available to download on SSRN. It builds on many of the themes developed on my posts here ― the rejection of judicial deference on constitutional issues, whether to legislatures or to the administrative state; the imperative to renounce the use of the Charter‘s “notwithstanding clause”; and the perils of “living constitutionalism”. Some of these, notably the issue of deference to administrative interpretations of constitutional law and constitutional interpretation, I will also be pursuing in future work. (Indeed, the first of these is the subject of the paper I will be presenting at the Journal of Commonwealth Law symposium next month.)

I am very grateful to Ms. Baron and Professor St-Hilaire for having given me the opportunity to present these thoughts, and write them up for publication. I am also grateful to Justice Bradley Miller, of the Court of Appeal for Ontario, who gave me thoughtful comments when I presented my chapter (then still very much in draft form) at the 2018 Runnymede Society conference, as well as to Kerry Sun, who was a very helpful editor. And I am looking forward to reading the other contributions in the volume, once I am done preparing the talks I am about to give in the coming weeks.

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.

In the Beginning

Learning about, and from, Pierre Trudeau’s 1968 proposal for what would become the Canadian Charter

Canadian judges and lawyers, including of the academic variety, tend not to think very highly of our constitutional history. This is, in part, because we ― and I must include myself in this ― do not know it as well as we should. There is an unhealthy feedback loop at work: a predisposition to be dismissive of the past fosters ignorance about it, which in turn makes it easier to be dismissive. The good news is that, once one starts looking into this history about which we have so much to find out, it is easy to find fascinating stories to learn, and to learn from.

Case in point: the proposal for “A Canadian Charter of Human Rights“, put forward in early 1968 by then-Justice Minister Pierre Trudeau, made available by the wonderful resource that is the Primary Documents project. I have to admit: I didn’t really know anything about this text before coming across it recently. But it is, surely, of considerable interest, if we accept that ― like every other rights-protecting text from the Magna Carta onwards ― the Canadian Charter of Rights and Freedoms, the direct descendant of the 1968 proposal, isn’t just a shadow in Plato’s cave, but a document that was written and enacted by particular people, at a particular time, in a particular place. And in addition to both the interest that I think this text deserves and the way in which it illustrates the value of constitutional history more broadly, it also sheds some light on ongoing debates.


Trudeau began his introductory chapter by claiming that “Interest in human rights is as old as civilization itself.” (9) (This, I am afraid, is reminiscent of bad student work.) For a long time, he wrote,

these rights were known as ‘natural’ rights; rights to which all men were entitled because they are endowed with a moral and rational nature. … These natural rights were the origins of the western world’s more modern concepts of individual freedom and equality. (9)

Trudeau appealed to Cicero and Aquinas, as well as Locke and Rousseau, and quotes at some length from the Declaration of Independence. I’m not sure that his presentation of the concept of natural rights is fully accurate, but his reliance on these authorities as the starting point of an argument for constitutional protection of human rights is relevant to the recent debates about the nature and origin of the rights protected by the Charter.

Another point which has been the subject of recent discussion that Trudeau’s introduction addressed was that of Parliamentary supremacy. Trudeau was quite clear that his proposal involved “some restriction on the theory of legislative supremacy”, although this theory, he said, “is seldom pressed to its full extent”. (11) Equally clear, as will appear below, was his understanding that the courts would have the last word on the meaning and import of the rights guarantees that he proposed adding to the Constitution. The point of the exercise was to secure “the fundamental freedoms of the individual from interference, whether federal or provincial”, and also to “establish that all Canadians, in every part of Canada, have equal rights”. (11)


This theme of inviting judicial enforcement of rights’ guarantees is further developed in the next chapter. Trudeau discusses the Canadian Bill of Rights, and finds it wanting because it is “not a constitutional limitation on Parliament, only an influence”, (13) and has not been vigorously enforced by the courts. Even if it had been, it would, like provincial legislation protecting human rights, be subject to repeal through the ordinary legislative process. In short,

a constitutionally entrenched Bill of Rights is required which will declare invalid any existing or future statute in conflict with it. Language in this form would possess a degree of permanence and would over-ride even unambiguous legislation purporting to violate the protected rights. (14)


The next Chapter outlines the contents of the proposed “charter of human rights”. It explains how existing law deals with each right it proposes to protect ― what the existing protections, if any, are; how they are limited; and also how legislative powers affecting the right are distributed between Parliament and the provincial legislatures. In some cases at least, there is thought given to the wording of future constitutional clauses ― for example, “whether freedom of expression is best guaranteed in simple terms without qualification, or whether the limitations of this freedom ought to be specified” (16) ― which suggests that the Charter‘s text is not just a collection of “majestic generalities” that could just as easily have been cast in very different, if equally general, terms. And there is a great deal of speculation about the way in which the courts will treat various rights, if they are constitutionally entrenched. This speculation is informed by references to Canadian case law, where it exists, as well precedents from the United States. There are also occasional references to the European Convention on Human Rights.

Some future controversies are already foreshadowed in Trudeau’s discussion. For example, the section on the freedom of religion highlights “the imposition of Sunday closing of businesses on Christians and non-Christians alike” ― which would, indeed, produce one of the first Supreme Court decisions based on the Charter, R v Big M Drug Mart Ltd, [1985] 1 SCR 295. For its part, the section on “life, liberty and property” ― note that, as Dwight Newman and Lorelle Binnion have pointed out, Trudeau was quite keen on entrenching some form of constitutional protection for property rights ― anticipates the issue in another early Charter case, Re BC Motor Vehicle Act, [1985] 2 SCR 486 about whether substantive or only procedural constraints exist on deprivations of “life, liberty and the security of the person”. In 1968, Trudeau thought, based on the jurisprudence under the “due process clause” of the Fifth and Fourteenth Amendments to the U.S. Constitution

that the guarantee [of due process] as applied to protection of “life” and personal “liberty” has been generally satisfactory, whereas substantive due process as applied to “liberty” of contract and to “property” has created the most controversy. It might therefore be possible to apply the due process guarantee only to “life”, personal “liberty” and “security of the person”. The specific guarantees of procedural fairness set out elsewhere in the bill would continue to apply to any interference with contracts or property. In this fashion the possibility of any substantive “due process” problems would be avoided. (20)

Of course, the example of property rights shows that what was ultimately enacted in 1982 was not always what Trudeau had wanted in 1968. Still, given the widespread conviction that the Supreme Court’s holding in the Motor Vehicle Act Reference that the “principles of fundamental justice” which must be respected when depriving a person of “life, liberty or security of the person” were not only procedural but substantive too went against with the wishes and expectations of the Charter‘s framers, it is interesting to note that the Supreme Court’s interpretation is actually quite consistent with Trudeau’s original proposal.

There are instances, admittedly, where Trudeau’s powers of prediction failed. For example, he wrote that “a court would likely be extremely reluctant to substitute its opinion of a proper punishment for that of the legislature”. (21) Stephen Harper, not to mention Justice François Huot of the Québec Superior Court, might have a thing or two to say about that. Trudeau thought that constitutionalizing the presumption of innocence would not mean “that the various federal and provincial penal statutes which contain ‘reverse onus’ clauses … will be declared unconstitutional”. But of course it was just such statute that was in fact declared unconstitutional in R v Oakes, [1986] 1 SCR 103.

And property rights weren’t the only ones that he thought important but the Charter ended up not protecting: so was the right to a fair hearing in civil and administrative proceedings. On the other hand, some rights that Trudeau did not think advisable to incorporate in the constitution were read into it by judicial fiat. Thus, notably, Trudeau listed “the right to form and join trade unions” along with other rights “which seek to ensure some advantage to the individual and which require positive action by the state”, (27) and which should not be protected by his proposed “charter of human rights”. That is because “[i]t might take considerable time to reach agreement on the rights [in this category] to be guaranteed and on the feasibility of implementation”. (27) Someone should have told the Supreme Court before it decided in Health Services and Support – Facilities Subsector Bargaining Assn v British Columbia, 2007 SCC 27, [2007] 2 SCR 391, that a right to collective bargaining would have been “within the contemplation of the framers of the Charter“. [78]


The very brief final chapter in Trudeau’s text suggests that egalitarian and linguistic rights might have to be implemented gradually, after political and legal rights have been protected, and muses on the advisability of special provisions for wartime and other emergencies. Section 32(2) of the Charter, which provided that equality rights would only come into effect three years after the rest of the Charter, seems to reflect the former concern, as does, in part, section 59 of the Constitution Act, 1982 which requires Québec’s consent ― which has never been given ― for the application of section 23(1)(a) to the province. No special provision has been made specifically to accommodate the concern about emergencies, though Trudeau actually contemplated the possibility of leaving it to “the courts to determine what limitations are made necessary in times of crisis”. (30)


In case I have not made this sufficiently clear already: these are only one man’s ideas about what a future constitutional charter of rights for Canada should look like and accomplish. To be sure, the man was influential ― indeed his influence was decisive in Canada having a constitutional charter of rights 14 years later ― and the ideas were given the stamp of approval by the government of which he was part. But many years would pass, and many governments would change, before these ideas would become law, and then, as noted above, only in a much modified form.

It is the law that was enacted that binds Canadian governments, and Canadian courts. As I have unfortunately had occasion to note here, Pierre Trudeau’s political programme is not the appropriate object of constitutional interpretation, “and the courts’ duty is to apply the Charter as it has been enacted, and not to expand it forever until the day the just society arrives”. For the better and for the worse ― often much for the worse ― the ideas of other political actors and members of the civil society helped shape Charter as it developed from a political proposal to a constitutional law.

Nevertheless, the original proposal of which the Charter is the consequence deserves our attention. Although in no way binding or definitive, it sheds some light on important controversies surrounding the Charter, some of which are ongoing to this day ― in part, I would argue, because we have not paid sufficient attention to history. Studying this history is a way not only of indulging our curiosity ― though there’s nothing wrong with that ― but also of reminding ourselves that the Charter, and our constitution more broadly, was the product of specific circumstances and ideas. For all their flaws, these circumstances and ideas were more interesting and praiseworthy than those who denigrate them in order to make the constitution that they produced into a blank canvas onto which their own preferences can be transposed care to admit. The constitution is neither such a blank canvas nor a projection from a Platonic world of forms that must be interpreted by philosopher kings in judicial robes for our edification and government. It is a law, and must be interpreted as such.

“Clear Enough”

Some thoughts on statutory interpretation.

As I finish my graduate studies at  Chicago, it struck me that a major theme of legal design is the degree of perfection (if any) we should expect from legal rules. Drafted legal rules—whether by the legislature or judiciary—will always be over and underbroad, because rules of general application cannot foresee every idiosyncratic individual application. In such a case, the extent to which a perfect rule can be created is dependent on the extent to which we balance the error rate of application with the ease of administrability of a straightforward rule. Here, we will never come to a perfect balance, but we can try to come to something that is defensible and workable.

The same sort of consideration applies in the field of statutory interpretation. The most important issue in statutory interpretation is the clarity exercise—how clear is clear enough? Finding that a statutory text is clear on its face leads to a number of important consequences. For one, the Supreme Court has said that where text is “precise and unequivocal, the ordinary meaning of the words play a dominant role in the interpretive process” as opposed to purpose (Canada Trustco, at para 10). Additionally, the use of Charter values in statutory interpretation to gap-fill only arises where there is ambiguity in the ordinary textual meaning (BellExpressVu, at para 28). And, as Gib Van Ert points out, the Federal Court of Appeal seems to be adopting a similar rule in the context of international law.

Some may object at the outset to a consideration of “clarity” as a means of discerning legislative intent on a particular subject. This line of opposition is deeply rooted in the idea of legal realism, with its skepticism of judicial modes of reasoning and the rejection of abstract legal thought as a means to come to clear answers on the law. Representative works in this regard include John Willis’ “Statutory Interpretation in a Nutshell,” where he argues that, in modern legislation which uses wide language (often to delegate authority to others), literal interpretation does no good, essentially because the language is broad and unclear. And he notes that even if interpretation could be  clear or plain on its face, there are differences between judges as to what “plain” constitutes (see 10 and 11). Additionally, Karl Llewellyn’s classic article on the “dueling canons of interpretation” sheds doubt on the use of the canons of statutory interpretation to come to any clear meaning that is not inspired by motivated reasoning. Underlying each of these important critiques is a belief in the relativism and contingency of language. Clarity, on this account, is probably a fool’s errand, in part because ascribing an intent to the legislature is difficult with open-textured language, and in part because language itself is inherently unclear. If this is true, it will be the rare case indeed where a court should be convinced that a text is clear.

While this might sound good to a lawyer’s ear—especially a lawyer that is paid money to exploit ambiguities—it does not comport with the way we use language in the majority of cases. And this is where the example of crafting legal rules comes into handy. One might wish to craft a legal rule to cover all of the interstitial, idiosyncratic applications—ones that are weird or abnormal. But then we create a rule that might work well in the individual case, and not in the general run of cases. Instead, we should craft legal rules based on the 98% of cases, not the 2%: see Richard Epstein’s Simple Rules for a Complex World on this score. In the realm of statutory interpretation, this means that we should start with the going-in, commonsense presumption that language is generally clear in the majority of circumstances, after a bit of listening and synthesis. People transact in the English language everyday with no major kerfluffles, and even conduct complex business and legal dealings without requiring a court to opine on the language they are using. This underlying mass of cases never makes it to court precisely because English works. The problem with statutory interpretation cases, then, is the major selection effect they present. The cases that make it to court, where the rules are developed, are the cases that are most bizarre or that raise the most technical questions. Those are not the cases on which we should base rules of general application. Instead, the rule should simply be that English works in most circumstances, as evidenced by the fact that each of us can generally communicate—with only small hiccups—in the day-to-day world.

If that is the rule adopted, and if legal language is really no different in kind (only in degree of specificity and technicality), then a court should not be exacting in its determination of the clarity of a statutory provision. That is, if language generally works on first impression, then there is no need for a court to adopt a presumption that it doesn’t work, and hence that something greater than “clear enough” is required to definitively elucidate the meaning of a text. We should merely assume that language probably works, that legislatures know language, and that courts have the tools to discern that language. While we should not assume that language is perfect, we should at least assume that it is workable in an ordinary meaning sense.

This approach also has the benefit of commonsense. Perfection is not of this world.  The legal realists put way too high a standard on the clarity of language, to something approaching perfect linguistic clarity rather than semantic workability. We should not craft legal rules around the fact that, in some far-off circumstances, we can imagine language not working.

What does this mean in operation? The American debate over Chevron deference supplies a good example. Chevron holds that where Congress has spoken to the precise question at issue, courts should not afford deference to an agency’s interpretation of law. This is Chevron Step One. If Congress has not spoken clearly, the court moves to Chevron Step Two, where it will defer to the interpretation and uphold it if it constitutes a reasonable interpretation of law. In a recent case, Justice Gorsuch concliuded at Chevron Step One that the text was “clear enough,” so that deference should not be afforded. The clear enough formulation is reminiscent of Justice Kavanaugh’s article, where he explains the various divisions among judges about clarity:

I tend to be a judge who finds clarity more readily than some of my colleagues but perhaps a little less readily than others. In practice, I probably apply something approaching a 65-35 rule. In other words, if the interpretation is at least 65-35 clear, then I will call it clear and reject reliance on ambiguity-dependent canons. I think a few of my colleagues apply more of a 90-10 rule, at least in certain cases. Only if the proffered interpretation is at least 90-10 clear will they call it clear. By contrast, I have other colleagues who appear to apply a 55-45 rule. If the statute is at least 55-45 clear, that’s good enough to call it clear.

Kavanaugh’s approach is probably closer to the right one, if we accept the general proposition that language will be workable in the majority of cases. If there is no reason to doubt language, then clarity will be easier to come by. It is only if we go in assuming the case of unworkability that clarity becomes a fool’s errand. But from a perspective of legal design, this is not desirable.

Law has a reputation for being a highly technical field, with a laser focus on commas, semicolons, and correcting the passive voice. But at the level of designing legal rules, including rules governing language, the best we can hope for is workability, not technical precision. This is because designing rules involves tradeoffs between incentives, administrability, and fit. And because humans are not perfect, we cannot design rules at this level of abstraction that are perfect. As a result, in the language context, the best we can and should do is workability in the general run of cases.

Concurring Opinion

Does the Charter’s “notwithstanding clause” exclude judicial review of legislation? Not quite!

Earlier this month, Grégoire Webber, Eric Mendelsohn, and Robert Leckey published an interesting challenge to what they termed “[t]he faulty received wisdom around the notwithstanding clause” over at Policy Options. Professor Webber, Mr. Mendelsohn, and Dean Leckey argue that the invocation of section 33 of the Canadian Charter of Rights and Freedoms, the notorious “notwithstanding clause”, by a legislature that enacts a statute does not fully insulate that statute against judicial review. Only the consequences of such review, not its availability, are affected. A court can still declare a statute protected by the “notwithstanding clause” to be contrary to the Charter ― albeit that the statute will continue to apply. This is an intriguing argument, and I think that it is correct.

Section 33(2) of the Charter provides that “[a]n Act or a provision of an Act in respect of which a declaration made under this section is in effect shall have such operation as it would have but for the provision of this Charter referred to in the declaration.” Professor Webber, Mr. Mendelsohn, and Dean Leckey point out that “The word ‘override'”, often used to describe section 33, “appears nowhere and there is no mention of ‘judicial review’. Rather, the text of section 33 focuses on shielding a law’s ‘operation’.” It excludes the application section 52(1) of the Constitution Act, 1982, which would normally render a provision or statute inconsistent with the Charter “of not force or effect to the extent of the inconsistency”. But this does not prevent a court from declaring that an inconsistency exists in the first place.

I agree, and would add a further textual point. Section 33(1) authorizes the enactment of legislation that will “operate notwithstanding a provision included in section 2 or sections 7 to 15 of this Charter”. One provision that is not subject to section 33 is section 24, the Charter‘s internal remedial provision. Pursuant to section 24(1),

[a]nyone whose rights or freedoms, as guaranteed by this Charter, have been infringed or denied may apply to a court of competent jurisdiction to obtain such remedy as the court considers appropriate and just in the circumstances.

Normally, if one’s rights are infringed by legislation, the “remedy that is appropriate and just in the circumstances” is a declaration of invalidity pursuant to section 52(1). The invocation of section 33 of the Charter changes “the circumstances”, however, so that ― for as long as it applies ― it is no longer constitutionally “appropriate” for a court to issue a remedy that affects the “operation” of the statute protected by the “notwithstanding clause”. But it would be wrong to make the leap from that incontrovertible truth to the much broader ― and textually unsupported ― proposition that no judicial remedy is “appropriate … in the circumstances” that include an operating “notwithstanding clause”. Rather, a court faced with a challenge to a statute protected by the “notwithstanding clause” must still strive to issue a “just” remedy within the constraints of section 33; that is to say, a remedy that addresses the violation of claimant’s rights (if any) without purporting to affect the operation of the statute.

As Professor Webber, Mr. Mendelsohn, and Dean Leckey suggest, a bare declaration of inconsistency, which does not purport to render the inconsistent statute “of no force or effect”, would seem to be a remedy that is (however minimally) just, and constitutionally appropriate in circumstances that include an operating “notwithstanding clause”. As they note, the New Zealand Supreme Court recently came to a similar conclusion in Attorney-General v Taylor, [2018] NZSC 104. In Taylor (about which I wrote here), the majority held that a declaration of inconsistency was an appropriate remedy that can serve to vindicate the rights affirmed by the New Zealand Bill of Rights Act 1990 within the constraints imposed by section 4 of that Act, which prevents the courts from invalidating or refusing to apply inconsistent legislation. Even when no particular consequence flows from the declaration, it is still of value to the claimant, and granting it is in keeping with the courts’ role of saying what the law is.

This point is particularly apposite in the Canadian context, since the Charter ― even when section 33 is invoked ― is part of what section 52 of the Constitution Act, 1982 describes as “the supreme law of Canada”. As Professor Webber, Mr. Mendelsohn, and Dean Leckey point out, the courts have always stressed their responsibility for setting out the meaning of this law (well, always except when they follow Doré v Barreau du Québec, 2012 SCC 12, [2012] 1 SCR 395). This is so even in cases where, for one reason or another, the courts consider that their remedial powers do not reach as far as their power to articulate the law. Professor Webber, Mr. Mendelsohn, and Dean Leckey mention Canada (Prime Minister) v Khadr, 2010 SCC 3, [2010] 1 SCR 44, which is one such case; Reference re Secession of Quebec, [1998] 2 SCR 217 is another well-known example. The Canadian constitutional framework, even more than the New Zealand’s, is different from the Australian one, where the High Court held, in Momcilovic v The Queen, [2011] HCA 34, that the making of bare declarations of inconsistency was not a judicial function or even incidental to a judicial function, and so not something that the courts could constitutionally be asked to do.

Another point worth taking away from Taylor is that declarations of inconsistency should not be regarded as addressed to the legislature. Rather, they are vehicles by which the courts point out that the legislature has abused its powers, and the courts are prevented to do more about that fact than simply acknowledge it. The courts should not be thinking in terms of a dialogue with the legislature; it doesn’t matter whether the legislature is of a mind to take the courts’ judgment seriously. Professor Webber, Mr. Mendelsohn, and Dean Leckey suggest that ,”[i]nformed by the reasoned, evidence-based judgment of an impartial, independent court, the government might amend its policy or decide to allow section 33’s protection to lapse”. I suspect that this is a too optimistic ― certainly the New Zealand Parliament appears to be in no mind to remedy the inconsistency with the Bill of Rights Act identified in Taylor (which concerned the disenfranchisement of prisoners serving short sentences). But this doesn’t matter. It is the courts’ duty to say what the law ― and a fortiori the supreme law ― is, Parliamentary indifference be damned.

Professor Webber, Mr. Mendelsohn, and Dean Leckey’s argument that the invocation of section 33 of the Charter does not exclude judicial review, but only limits the consequences that can result from such review is novel, but I think that it is correct. They are right that, by its terms and within its constitutional context, “[s]ection 33 secures a law’s operation; it does not open a Charter black hole”. Given the Canadian provinces’ newfound penchant for relying on section 33, which I fear is only the start of a sinister trend, we may well soon find out what the courts will make of their idea.

I Said Don’t Do It

The federal government is wrong to involve Québec in the process of appointing the next Supreme Court judge

In 2014, after the Supreme Court invalidated the appointment of Justice Nadon to one of its seats reserved for Québec judges or lawyers, the federal government got the Québec government to propose a shortlist of candidates for the vacant-again position. This process resulted in the appointment of Justice Gascon to the Supreme Court. The federal government meant the outsourcing of the shortlist to be a one-off; the Québec government was hoping that it would create a precedent. Québec’s wishes were ignored when the next appointment to one its seats (that of Justice Côté) was made.

But now Justice Gascon is now retiring ― sadly, much before his time ― and a version of the process that produced his appointment is being brought back. As the Canadian Press reports,

[t]he federal and Quebec governments have reached what the province is calling a historic deal that ensures it will play an active role in the process of selecting the next Supreme Court of Canada justice from Quebec.

An advisory committee similar to those used for previous appointments made by the current federal government submit will then

submit a shortlist of candidates to the federal and provincial justice ministers. … [T]he premier of Quebec will also provide an opinion and forward a recommendation to the prime minister, who will make the final decision weighing the recommendation of the federal justice minister and Quebec’s input.

The provincial government’s role is, if I understand correctly, not as important as in the 2014 process, since it doesn’t extend to unilaterally determining the Prime Minister’s range of choices. But it is still significant. The province seems delighted. The Canadian Press writes that the provincial justice minister “called the deal precedent-setting” ― yes, again ― “saying it would allow the province to take a ‘direct and significant part’ in the judicial appointment”.

The rest of us should not be happy. In fact, we should be rather angry. I criticized the 2014 process at some length here, and I believe that that criticism is still applicable, albeit in a slightly watered-down form, to the new process. It is common enough for members of the Canadian chattering classes to claim that the federal government’s power of appointing Supreme Court judges without taking provincial preferences into account is a defect in our federal system. But this view is mistaken. Here’s part what I said in 2014 (with references updates):

[H]ow much of a flaw is it really that the federal government appoints judges unilaterally? In practice, the Supreme Court’s recent blockbuster decisions ― the one concerning the eligibility of Justice Nadon, Reference re Supreme Court Act, ss. 5 and 6, 2014 SCC 21, [2014] 1 SCR 433 and that in the Reference re Senate Reform, 2014 SCC 32, [2014] 1 SCR 704 ―, as well as Reference re Securities Act, 2011 SCC 66, [2011] 3 SCR 837, which declared a proposed federal securities regulator unconstitutional belie any claim that the Supreme Court is biased in favour of the federal government.

And even at the level of theory, there is a good argument to be made for unilateral federal appointments. Canadian history has borne out James Madison’s famous argument in Federalist No. 10 that small polities are more vulnerable to “faction” and the tyranny of the majority than larger ones. Our federal governments have tended to be more moderate than provincial ones, and less susceptible to takeovers by ideological entrepreneurs from outside the Canadian mainstream, whether the Social Credit of Alberta or the separatists of Québec. Foreseeing this, the framers of the Constitution Act, 1867 gave the power of appointing judges of provincial superior courts to the federal rather than the provincial governments. It stands to reason that the judges of the Supreme Court, whose decisions have effect not only in one province, but throughout Canada, should a fortiori be appointed by the government more likely to be moderate and representative of the diversity of the views of the country ― that is to say, by the federal government.

Québec’s case is illustrative. The federal government presumably is comfortable with, or at least not very worried about, outsourcing the selection of potential Supreme Court judges to a relatively friendly, federalist government. Would it have felt the same way if the Parti Québécois ― not only separatist, but also committed to the infamous “Charter of Québec Values” (which the federal government had vowed to fight in court!) had won the recent provincial election? 

The latest developments sure give us some food for thought on this last question. The Parti Québécois, it is true, not only remains out of government, but is currently the fourth-largest party in Québec’s legislature. Yet its idea of purging the province’s public service of overtly religious persons ― especially if they are overtly religious in a non-Catholic way ― is alive, kicking, and in the process of being enacted into law, as Bill 21, by the Coalition Avenir Québec’s government. This is the same government, of course, that its federal counterpart wants to involve in the appointment of the judges who may yet be called upon to pronounce on Bill 21’s consistency with the constitution.

Back in the sunny days of 2015, when illusions about the current federal government being formed by the “Charter party” were still possible, the Prime Minister wrote the following to his Attorney-General:

[Y]our overarching goal will be to ensure our legislation meets the highest standards of equity, fairness and respect for the rule of law. I expect you to ensure that our initiatives respect the Constitution of Canada, court decisions, and are in keeping with our proudest legal traditions. You are expected to ensure that the rights of Canadians are protected, that our work demonstrates the greatest possible commitment to respecting the Charter of Rights and Freedoms, and that our government seeks to fulfill our policy goals with the least interference with the rights and privacy of Canadians as possible.

The “Mandate Letter” in which these wonderful commitments are set out is still on the Prime Minister’s website, although its original addressee was eventualy fired for acting like an actual Law Officer of the Crown and not a political weather-wane. But the same Prime Minister’s government is now going out of its way to hand over part of its constitutional responsibility for appointing the judges of Canada’s highest court to a provincial government bent not only on trampling on fundamental freedoms, but also on insulating its actions from review for compliance with the Charter. I should have thought that this is an odd way of respecting the Constitution of Canada, of ensuring that the rights of Canadians are protected, and of demonstrating the greatest possible commitment to respecting the Charter of Rights and Freedoms. But what do I know?

Well, I know this. Five years ago wrote that

[t]he power to appoint Supreme Court judges belongs to the federal government, and it alone, for good reason. … [T]he constitutional edifice built in 1867 (and 1875, when the Court was created, and then 1982 when it was, so it says, constitutionally entrenched) has weathered some great storms, and given us all shelter and comfort. It is in no danger of crumbling. Do not try to rebuild it.

Don’t do it. Just don’t.

Judicial Supremacy, Again

Another attack on judicial supremacy misses the mark

Last week, the Québec government put forward a bill that will, under cover of the Canadian Charter’s “notwithstanding clause” and its provincial analogue, declare irreligion the province’s official creed and bar a multitude of office-holders and public employees from wearing religious symbols. Just a couple of days before, over at Policy Options, Brian Bird published the latest contribution to the judicial-supremacy-bad-legislatures-good genre that has been undergoing something of a revival in Canada of late. It is, alas, no more compelling than all the others.

Mr Bird beings by asking two questions: “Is leaving this responsibility [for upholding the constitution] solely in judicial hands the best way of upholding the supreme law of a liberal democracy such as Canada? Does our Constitution even call for judicial supremacy in constitutional matters?” The first question is misdirection. No one, to my knowledge, has suggested that, since the courts are able to enforce the constitution, the other branches of government should ignore it. The answer to the second question, as I have argued here, is a resounding “yes”.


Let me start with that second question. (A fuller statement of my views on it is in the post linked to in the previous paragraph.) Mr. Bird claims that section 52(1) of the Constitution Act, 1982, which provides that “[t]he Constitution of Canada is the supreme law of Canada, and any law that is inconsistent with the provisions of the Constitution is, to the extent of the inconsistency, of no force or effect”, “does not identify courts as the sole or final arbiters of constitutionality. It identifies no particular branch of the state as uniquely responsible for these tasks.” That much is true: section 52(1) does not explicitly mention the courts. But that’s because it doesn’t have to.

As Mr. Bird himself helpfully explained elsewhere, section 52(1) was not an innovation in the Canadian constitutional system, but rather a replacement for the Colonial Laws Validity Act 1865, which required the courts to uphold the supremacy of imperial statutes, including what we now call the various Constitution Acts. If the framers of the Constitution Act, 1982 had wanted to deny the judiciary this authority, they would have given us some indication of the fact, instead of remaining cryptically silent. Far from doing so, the framers also sought to entrench the Supreme Court in the constitution ― or so the Supreme Court itself has told us. Why in the world would they have done that if they didn’t think that the court had a special responsibility for enforcing the constitution itself?

And there is more, as I pointed out in the post linked to above:

[T]he phrase “supreme law” (emphasis added) [in section 52(1) also suggests that, like any law, the Constitution of Canada is subject to interpretation and enforcement by the courts―not by legislatures. Granted, by 1982, the Supreme Court had conceded deference on the interpretation of some legal provisions to administrative adjudicators. But that concession was premised―wisely or not is beside the point here―on these adjudicators’ expertise, including legal expertise in their particular area of jurisdiction. I do not think that Parliament would have been understood to have such expertise.

Mr. Bird, for his part, suggests that “the Constitution’s status as the ‘supreme law’ … demands compliance with the Constitution, not a particular mechanism for enforcing compliance.” Yet the normal mechanism for enforcing compliance with law is adjudication, and even to the extent that enforcement can be delegated to non-judicial institutions (and, to repeat, these are supposedly expert institutions specialized in administering specific areas of the law) the courts retain a power of review over their work. If the 1982 framers contemplated some other mechanism for ensuring compliance with the law they were enacting, they would undoubtedly have said so. In short, in my view the original public meaning of section 52(1) ― in the context of its predecessor provision’s text and history ― clearly requires “judicial supremacy in constitutional matters”.


Coming back, then, to Mr. Bird’s first question, whether we would not be better off if all branches of government, and not just the courts, were engaged in upholding the constitution, one can only say, “of course we would”. Mr. Bird does not identify anyone who might disagree but, for the record, I support his view that “[l]egislatures should repeal unconstitutional laws”. I have misgivings about Mr. Bird’s suggestion that “the executive should not enforce” laws it deems unconstitutional, partly for positive law reasons co-blogger Mark Mancini raises in his latest post, and partly for philosophical reasons I refer to here. But the point is a difficult one, and Mr. Bird may well be right. And of course both legislatures and the executives, so far as the law allows, are free to, and should, do more to uphold the constitution than the courts will let them get away with.

The real question, however, is not whether it would be desirable for Canadian legislatures and executives to endeavour to enforce the constitution, but whether they are at all likely to do so. The answer, sadly, is that they are not. While it is true, Mr. Bird notes, that “[g]overnment lawyers frequently give opinions on the constitutionality of proposed legislation [and] [i]n some cases … have a statutory duty to do so”, the standard they apply for concluding that proposed legislation is constitutional is ridiculously low. (It is close, in effect, to a puke test, or to asking whether a colleague defending the statute would be laughed out of court.) And, as I have noted here, when politicians are required to make their own constitutional judgments (in areas that are not justiciable), they “take the constitution no more seriously than when they act under adult judicial supervision. Actually, they do not care about it at all.”

This is not a uniquely Canadian affliction, of course. In New Zealand, successive Attorneys-General have applied a higher standard than their Canadian colleagues to concluding that a proposed enactment would infringe the Bill of Rights Act 1990, but their not infrequent reports to this effect have largely been ignored by Parliament. And even when the courts have pointed out inconsistencies between ordinary legislation and the Bill of Rights Act, contrary to Mr. Bird optimistic prediction, these indications have not “influence[d] the deliberations of governments and … foster[ed] dialogue between branches of the state on constitutional issues”. Legislation flatly contrary to the Bill of Rights Act remains on the books unaltered.


The attack on judicial supremacy and attempts to discredit the judiciary as constitutional enforcer tend, ultimately, to be based on unwarranted optimism about the interest of the “political branches” for the constitution. In my view, there is little cause for such hopefulness. It is true that jurisdictions with judicially enforceable constitutions, such as New Zealand, may remain fairly free ― though it is also true that New Zealand is vulnerable to illiberal policy shifts against some of which a supreme constitution might offer a modicum of protection. But there is nothing to be gained, and likely something to be lost, by giving up on judicial enforcement of supreme constitutional law.

The revival of arguments in favour of this option, coinciding as it does with a shameless political trampling on constitutional constraints and rights illustrated by Québec’s anti-religious legislation, is puzzling and counter-productive. The courts, of course, are very far from perfect in their capacity as constitutional enforcers. But we should be insisting that they become better at this job, not suggesting that they might as give it up.