Tag: proportionality analysis

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.

Doing Right on Rights

Why the Supreme Court was right to find the disenfranchisement of Canadians abroad unconstitutional

In my last post, I summarized the majority, concurring, and dissenting opinions delivered in Frank v Canada (Attorney General), 2019 SCC 1, which held that the denial of the franchise to Canadians who have resided outside the country for more than five years is unconstitutional. As noted there, I believe that the majority, whose opinion was delivered by Chief Justice Wagner was correct so to hold. In this post, I explain why, and also make some observations about the strongly-worded dissent by Justices Côté and Brown. To make my biases clear once again, I remind readers that I am a Canadian abroad myself, and have been for six of the past eight years. While I vote rarely and reluctantly, I am emotionally invested in the issue of whether my right to do so can be taken away. Feel free to discount the following accordingly.

While I am the subject of emotions, let me say this. Justices Côté and Brown assert that the denial of the right to vote to Canadians abroad “is not a distinction based on moral worth”. [168] By my lights, that’s precisely what it is. On their own view, it is the product of “normative conceptions of what the Canadian political community is, and how it can best be protected and made to flourish”. [139] The conception embraced by Parliament in enacting the provisions invalidated in Frank sees the Canadian political community as excluding Canadians abroad, who are consequently less worthy of the franchise. Indeed, they are, according to the dissent, a threat from which the community must be protected, since it is “unfair to Canadian residents for their lawmakers to be elected by long-term non-residents who have no connection of any currency to their electoral district”. [153]

This is utterly wrong. As the Chief Justice rightly observes, many Canadians who live abroad maintain strong ties with Canada (and, I would add, often with the local community where they used to live and, in many cases, intend to return). As the Chief Justice also says, “[c]onversely, there may be citizens who have never left Canada but whose subjective commitment to the country is much weaker and who are less well versed in local issues”. [68] Indeed, though it would be impolitic for a judge to say so, “may be” in this sentence should read “obviously are”. As, for example, Ilya Somin and Bryan Caplan respectively have shown, voters are both ignorant and irrational ― rationally so, but ignorant and irrational all the same. Yet we would not generally accept disenfranchising voters on that basis; we do not inquire into the degree of connection a voter has with his or her local community, or the country, before issuing him or her a ballot paper. It is only, it seems, in the case of expatriates that these things actually matter. To me, this is strong evidence that what is at work here is not really a concern with the fairness or integrity of Canada’s electoral system, but a judgment, or rather prejudice, about the moral worth of those Canadians who are taken to have left the community, and must “rejoin[]” [153] before being allowed to take part in the community’s affairs.

Now, it is true that most Canadians abroad do not, in fact, go to the hassle and the expense of requesting a special ballot and returning it to Elections Canada. This means, of course, that the idea of non-resident voters swamping elections and deciding them at the expense of residents is far-fetched if not entirely implausible. But more importantly, as the Chief Justice points out, this means that those who do take the trouble value their involvement in Canada’s political life ― probably more so than a great many of their resident fellow-citizens. In the Chief Justice’s words, they “demonstrate[] a profound attachment to Canada”, [75] and it seems absurd to pretend otherwise on the basis of no evidence whatsoever.

This brings me to another issue: that of the correct approach to deciding whether legislation is unconstitutional because it unjustifiably contravenes the Canadian Charter of Rights and Freedoms. Justices Côté and Brown issue what they regard as a profound challenge to the way we not only discuss but also think about the relationship between Charter rights and policies that interfere with them. They note that it is commonplace to speak of such policies as “infringing” or even “violating” rights, only for these “infringements” or “violations” to be upheld, or “saved” by applying section 1 of the Charter. Yet, as they further point out, section 1 provides that the Charter “guarantees the rights and freedoms set out in it subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society” ― limits, not infringements or violations. So we should stop talking about justified infringements, and talk about limits instead. The majority is not interested, observing that the words “limits” and “infringements” have long been used interchangeably. (The Chief Justice is probably too polite to note this, but I am not: Justices Côté and Brown themselves spoke in terms of “infringement” as recently as seven months ago, in their excellent dissent in Law Society of British Columbia v Trinity Western University, 2018 SCC 32.)

I take the dissent’s textual point. The word section 1 uses is “limits”, and it arguably makes more sense to say that the state can justifiably limit rights, not that it can violate them. But I fail to see what great change to the way we actually think about rights and scrutinize the state’s possible interference with them follows from this. Contrary to some, indeed quite a few, of the participants in this blog’s recent 12 Days of Christmas symposium, Justices Côté and Brown show no interest in reconsidering the test for verifying the permissibility of limitations on rights that was first set out in R v Oakes, [1986] 1 SCR 103. They apply the same proportionality analysis, warts, subjectivity, and all, under the label of assessing the “limit” on the right to vote as the majority does when considering “infringement” of this right. They have much to say about the fact that the right to vote is a “positive” one, requiring legislation to “breathe[] life into” it, [142] but that it is true of many of the rights the Charter protects (notably, but by no means only, the various rights of criminal suspects and accused) and, more importantly, it simply does not follow that the legislation that “breathes life into the right” may not fail to do so to a constitutionally required standard. (Similarly, Justices Côté and Brown point out that the legislation imposing a five-year expiry period on expatriates’ franchise replaced that which gave no expatriates the vote. So what? The test of constitutionality is not whether Parliament comes closer to respecting the Charter than it once did, but whether it respects the Charter now.)

The real methodological disagreement between the dissent and the majority (as well as Justice Rowe’s concurrence) has to do with the level of deference each accords Parliament. The majority insists, at the outset, that “[a]ny limit on the right to vote must be carefully scrutinized and cannot be tolerated without a compelling justification”. [1] Although, as noted in yesterday’s post, it later wavers a little on this point, its approach is, indeed, one that refuses to take the government’s claims about the need for or usefulness of the legislation it examines on faith. By contrast, the dissent waxes deferential, “eschews rigid and technical application” [124] ― it’s not quite clear of what, but presumably of the justification requirements ―, and plays up “Parliament’s policy-making expertise”. [126]

Yet here the dissenters’ professed textualism fails them. To repeat, section 1 of the Charter says that “only such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society” (emphasis mine) can be countenanced. The constitutional text, therefore, requires a demonstration ― not judicial acquiescence on the basis that Parliament knows best. And this requirement, in turn, suggests that contrary to what Justices Côté and Brown say Charter rights can only be limited in order to deal with an identifiable problem, to address some specified mischief, and not merely because a legislature thinks that in an ideal political community these rights would be limited in this way. The enactment of the Charter, as a law superior to ordinary legislation, has taken that moral judgment out of the Canadian legislatures’ hands. (To be clear, this is a separate question from that of the permissibility of what used to be called “morals legislation”, which proscribes what the legislature sees as immoral behaviour, such as taking drugs. There is no argument that expatriates act immorally by voting.)

In short, by all means, let’s be careful with our language, and speak of reasonable limits on rights instead justified violations. But let’s also insist that limitations on rights, to qualify as reasonable, must be ones whose justification is capable of being demonstrated, rather than merely asserted, and is in fact demonstrated, rather than taken for granted. Indeed, I think that this substantive concern is rather more important than the semantic one. In Frank, it is the majority, not the dissent, that follows an approach that is closer to that required by the constitution.

In my view, the majority decided Frank more or less as it should have. It correctly insisted that any exclusion from the franchise except the one based on citizenship must be justified. Contrary to the dissent’s strident warnings, it does not follow that no exclusions could ever treated as reasonable limits ― only that the justification process for upholding these limits cannot be elided by saying that some limits on the right to vote must obviously be admissible. This goes even for the denial of the right to vote minors (which, as Ilya Somin, for example, has argued, is not as self-evidently reasonable as the dissenters would like us to think). I think that it would be quite easy to sustain the disenfranchisement of three-year-olds, the dissent’s scaremongering example, should anyone challenge it; but as for the denial of the vote to teenagers, I for one wouldn’t mind seeing the governments put through their justificatory paces.

The majority is also right to be skeptical of the government’s arguments based on an ill-defined “social contract” of which expatriates are allegedly no part. I’m not too impressed with the Chief Justice’s definition of Parliament’s objective in terms of “fairness” ― fairness is too capacious a word for my liking, and the dissent’s suggestion that Parliament was pursuing a currency of relationship between the voter and his or her community seems closer to the mark. But one should also acknowledge that objectives that are largely symbolic (or, as Justices Côté and Brown see it, moral) do not lend themselves to easy definition, and so inevitably compromise the quality of the Charter analysis. The majority’s skepticism about the existence of a rational connection between the objective of fairness and disenfranchisement of Canadians abroad is also warranted. Indeed, I would have liked the Chief Justice to have been bold enough to say that, given both the possibility that expatriates maintain current links with Canada and their Canadian communities, and the lack of any assessment of whether any other voters do so, disenfranchising expatriates alone based on their deemed lack of connection to Canada is actually irrational. But the Chief Justice is not the first judge not to want to go there.

I’ll live with that, because the Chief Justice ends up making the right arguments at the “minimal impairment” stage of the analysis. Like I had in criticizing the decision of the Court of Appeal for Ontario in this case, he points out that Canadians abroad can and do maintain close links with Canada; that they are affected by Canadian laws and government policies; that, moreover, “Parliament can change laws on its own initiative and thus alter the extent to which Canadian legislation applies to non-resident citizens”, [72] so that one cannot invoke the limited scope of current extra-territorial legislation to deny expatriates the vote ― I had called this “let[ting] the statutory tail wag the constitutional dog”; and that in any event “attempting to tailor Charter rights to the extent to which citizens are burdened, or not burdened, by Canadian laws would be an impossible exercise”. [71] If all this is not enough to qualify the disenfranchisement of expatriates as irrational, than it certainly suffices, as the Chief Justice says, to show that it is grossly overbroad, and not “minimally impairing” of the right to vote.

The Chief Justice’s reasons are not perfect. He does, as the dissent points out, get somewhat carried away in patriotic praise for the Canadian democracy, and there are shades of what I’ve been calling “constitutionalism from the cave” in his claim that “a broad interpretation of” the Charter‘s guarantee of the right to vote “enhances the quality of our democracy and strengthens the values on which our free and democratic state is premised”. [27] Constitutional interpretation must aim at ascertaining the text’s meaning, not at strengthening values or anything of the sort. Still, what the Chief Justice’s opinion for the majority does is substantially in line with the constitutional text, despite its rhetorical imperfections. (And still on the subject of rhetoric, or style: can someone please ask the Chief Justice and his colleagues to stop using the initialism “AGC”? It is a recent innovation ― the Court hadn’t done it at all before 2010 ― and not a good one. It is ugly and has a jargon-y feel that is quite at odds with the Chief Justice’s stated desire to make the Court’s work more accessible, including to laypersons.) But as majority opinions of Supreme Court have gone in the last few years, the one in Frank is up there with the best.

The Supreme Court has held that Canadians who live abroad cannot be disenfranchised based on, in effect, stereotypes about their lack of relationship with their home country. This is a relief. In doing so, it has insisted on carefully and critically examining the government’s claimed reasons for limiting Charter rights, and this is a good thing too, one that doesn’t happen often enough. There is something to learn from the dissenting opinion, too, about our constitutional vocabulary, and I hope that this lesson is not lost just because the substance of that opinion is bitterly disappointing. But the constitution was upheld in Frank, and so right was done. May 2019 bring more of that.

The Panglossian Peril

The dangers of naïve optimism in thinking about constitutional constraint

In a provocative paper recently posted on SSRN (and based on the HLA Hart Memorial Lecture delivered last year at Oxford), Frederick Schauer challenges a fairly common tendency to argue that apparent conflicts between rights and important interests, or among rights, are illusory, and that, properly understood, these rights and interests can be reconciled so as to avoid the conflict. Professor Schauer calls this tendency “Panglossianism”, after the obstinately and obliviously optimistic character of Voltaire’s Candide, and argues that it makes for muddled thinking that will end up compromising the rights that Panglossians purport to value. Professor Schauer makes important points, although I am not persuaded by his takeaway.

* * *

Professor Schauer wants us to recognize that we cannot have it all ― socially desirable policies fully implemented and rights fully protected at the same time. He laments

the common but nonetheless troubling tendency of many people to perceive (or distort) the empirical aspects of various interests in a manner that eliminates the conflict between them and other interests, or between those interests and the rights with which they may conflict. And although people sometimes thus perceive interests in ways that make rights appear cost-free, they also indulge in the equally common tendency to define rights in a way that similarly eliminates the constraints that rights sometimes impose on legitimate interests. (1-2)

To relate just one of the examples he uses, when it comes to prohibitions on hate speech, those who oppose them will often insist that hate speech is not especially harmful, or is not harmful in ways that anyone should really care about, so that upholding the right to freedom of expression has no real cost. Conversely, many of those who support the criminalization of hate speech invoke the mantra of “hate speech is not free speech”, similarly insisting that their preferred resolution of this issue is costless. “Panglossianism” can accordingly involve either a reading of the data (or speculation) about the effects of policies that minimizes their impact on rights, or a redefinition of rights or other constitutional rules that narrows them so as to ensure that a favoured policy is not precluded.

Professor Schauer argues that Panglossianism is a mechanism people deploy to deal with the threat of cognitive dissonance that they might experience if they acknowledge that their preferred policies and constitutional commitments are in tension, and even in conflict. It is easier to believe, and to say, that such conflicts are not real, or can easily avoided, than to deal with them, which would mean taking sides, recognizing that one is wrong about rights or that one’s preferred policy cannot be implemented.

Yet Panglossiansim is a problem, for two reasons. First, intellectual honesty requires us “to avoid attempting to see the empirical world through the rose-colored glasses of one’s own normative desires”. (18) And second, rights are safer if they are not justified on the basis of empirical claims that are either  weak to begin with, or at best “temporally or culturally contingent [so that] the right may turn out to be weakened or inapplicable under different empirical conditions”. (19) Rights are stronger if they are grounded in pure moral principle than if their continued existence depends on whether it is, or is not, relatively harmless. Indeed, if only harmless rights deserve protection, then not much is going to be protected at all. Professor Schauer warns that policy-makers (whether in the legislative or in the executive branch) can Panglossians too; they will tempted to insist that there is no conflict between their preferred policy and constitutional restrictions on their action. But

[u]nlike the rights-concerned Panglossian commentator or advocate who interprets – or distorts – the empirical data so as to eliminate a conflict between policies and rights, here the official is more likely to attempt to eliminate the conflict between policy preferences and constitutional constraints by understanding the constraints in a way that makes them inapplicable to the issue at hand. (22)

Panglossianism, Professor Schauer notes, can undermine not only rights protections, but all manner of intended constitutional rules. Resorting to it may be psychologically comforting, but it will weaken the very idea of constitutional constraints on governments’ pursuit of their preferred policies.

* * *

I think that Professor Schauer describes a real problem. It is indeed tempting to say that the enforcement of one’s favoured right does not compromise the attainment of valuable policy goals or the respect of other rights; it is similarly tempting to insist the implementation of one’s preferred policy conflicts with no real rights, properly understood. Debates about free speech are one area where this dynamic is especially visible, as Professor Schauer notes, but there are any number of others. It is arguable (which is not necessarily to say true) that the controversy over the federal government’s demand that religious groups “attest” to the compatibility between their “core mandate” and (some) Charter rights, about which I’ve written here, also involves Panglossian arguments on both sides.

And Professor Schauer is quite right to point out that Panglossianism can affect thinking about structural constitutional rules, and not just rights. Indeed, I would suggest that in Canadian constitutional law, Panglossianism is an especially strong danger in federalism jurisprudence. In Charter cases, section 1, which authorizes the imposition of “limits” to rights, channels the analysis into a more explicit consideration of the conflict between rights (which tend to be defined in broad and abstract terms) and policy reasons for restricting them. By contrast, the movement towards the erosion of the exclusivity of federal and provincial heads of power under the banner of “co-operative federalism” proceeds from the Supreme Court’s unwillingness to acknowledge the existence of conflict between what it sees as desirable policy and the federal division of powers. Federalism analysis simply makes no room for the acknowledgement of this conflict. This is not to say that we should change the way we approach it ― but we should beware the Panglossian dangers inherent in what we do.

Yet while I think that there is a great deal of truth to Professor Schauer’s diagnosis of the pathologies of Panglossianism, his prescriptions against it may not be especially salutary. Professor Schauer does not tell us much about how to assess what he sees as potentially-Panglossian claims about the effects of policies or the scope of rights. He warns against thinking, for instance, that not punishing hate speech is costless because such speech does not really cause any social evils. Fair enough ― those of us opposed to bans on hate speech on normative grounds will be tempted to downplay its effects. But what if it really doesn’t have any? Conversely, if hate speech really is socially harmful, that happens to align with the preferences of those who want to ban it. Both sides in this particular debate cannot, I think, be wrong at the same time. The mere fact that an empirical claim aligns with someone’s prior normative preferences cannot mean that the claim is wrong. The same applies to claims about the scope of rights (to the extent that these can be said to be correct or incorrect at all).

So while we should be wary of the dangers described by Professor Schauer, he has not convinced me to give up on empirical or otherwise contextualized thinking about rights in favour of a priori philosophizing. This is all the more so in the numerous cases that concern what might be described as marginal (possible) infringements of rights. Perhaps the hate speech question, which is about whether people can be prevented from saying certain things at all can be sufficiently resolved by an a priori insistence that such bans are never permissible. Note, though, that the argument wouldn’t work the other way: a case for banning hate speech can only be made if one is allowed to rely on empirical considerations (unless of course one takes the position that there is no right to free speech at all and anything can be banned). But what about, say, restrictions on financing political parties? Most people accept that at least some restrictions are acceptable (most people in North America, anyway; New Zealand has no limit on how much one can give, and seems to be doing just fine!). Many ― most, I hope ― would also agree that some restrictions are too extreme and cannot be justified. The issue is where to draw the line, and where to err in doubt. I don’t think that we can give remotely interesting answers to these questions without knowing something about the current practices of political fundraising and the likely effects of raising or lowering the existing restrictions. Again, Professor Schauer’s warnings about Panglossianism are relevant, but his suggestion that we resolve our questions by reference to first principles alone is not helpful.

Now, Professor Schauer is right, of course, that any empirically contingent answers might be inapplicable under different circumstances. He might be overstating the extent to which this is a problem: I’m not sure, for instance, that cultural contingency of rights protections is objectionable; it’s not obvious that rights must be the same everywhere and at all times. However, to the extent that, within a legal order, rights are implemented through judicially articulated constitutional doctrine, this doctrine risks being destabilized if the empirical or normative premises on which it is based are challenged by the evolution of society and of what we know about it. How to deal with this risk of instability (and its converse, the risk of a static doctrine divorced from reality) is a difficult question, to which I have no very good answers. But I doubt that we can avoid trying to get at some answers, at least, if only mediocre (and contingent!) ones.

Thinking about constitutional rules and their relationship with policy is a difficult business. Professor Schauer is right to remind us that we are too often tempted to oversimplify it by pretending that contradictions between our normative commitments and policy preferences are less significant than they really are. Unfortunately, he doesn’t offer us much by way of useful advice for identifying the exact situations where our thinking is so sidetracked, and his suggestion that we think more about abstract principle than about the real-world effect of policy does not strike me as especially helpful. Nevertheless, Professor Schauer’s warning is an important one, and we should heed it even if we conclude that we must continue exposing ourselves to the dangers he highlights.