Unmaking History

In the “free the beer” case, the Supreme Court shows ― again ― that it is the spoiled child of the Constitution

When it accepted to pronounce on the constitutionality of non-tariff barriers to inter-provincial trade, the Supreme Court had a chance to make history. In R v Comeau, 2018 SCC 15, the Court chose to unmake it instead. Far from “freeing the beer” and invalidating legislation that prevents bringing booze from one province to another and other regulatory schemes built on provincial protectionism, Comeau countenances even restrictions on inter-provincial trade that would previously have been thought flatly unconstitutional. In the process, it tramples over constitutional text and history, as well as logic.

Section 121 of the Constitution Act, 1867 provides that “[a]ll Articles of the Growth, Produce, or Manufacture of any one of the Provinces shall, from and after the Union, be admitted free into each of the other Provinces”. But free of what exactly? Of any and all regulation, or of just some particular kinds? In Gold Seal Ltd v Alberta (Attorney-General),  (1921) 62 SCR 424, the Supreme Court held that “free” meant “free from tariffs”. In Comeau, it was asked to revisit this holding. As the Court ― its members evaded responsibility for their (mis)judgment by attributing it to the institution, though I am looking forward to Peter McCormick or someone else exposing the true author(s) ― notes, this question is of the highest importance:

If to be “admitted free” is understood as a constitutional guarantee of free trade, the potential reach of s. 121 is vast. Agricultural supply management schemes, public health-driven prohibitions, environmental controls, and innumerable comparable regulatory measures that incidentally impede the passage of goods crossing provincial borders may be invalid. [3]

* * *

Before answering the interpretive question, however, the Supreme Court addresses a different one: whether the trial judge was entitled to depart from Gold Seal to hold that s. 121 applied to non-tariff barriers to inter-provincial trade. The judge had taken up the Supreme Court’s invitation, issued in Canada (Attorney General) v Bedford, 2013 SCC 72, [2013] 3 SCR 1101, to revisit precedent in light of newly available evidence. In Bedford and Carter v Canada (Attorney General), 2015 SCC 5, [2015] 1 SCR 331, which dealt with the constitutionality of the provisions of the Criminal Code relative to prostitution and assisted suicide respectively, the evidence that was held to allow lower courts to revisit Supreme Court precedent came mostly from the social sciences. In Comeau, the trial judge relied on new historical evidence about the context and original meaning of s. 121.

This, the Supreme Court insists, was not something that Bedford authorizes. Bedford “is not a general invitation to reconsider binding authority on the basis of any type of evidence”. [31; emphasis mine] What is required is a showing “the underlying social context that framed the original legal debate is profoundly altered”, [31] triggering the applicability of the Court’s “living tree” approach to the constitution. Historical evidence, which the court derides as “a description of historical information and one expert’s assessment of that information”, does not count: “a re-discovery or re-assessment of historical events is not evidence of social change”. [36]

In conversation with Maclean’s, Carissima Mathen said the Court “essentially chastised the trial judge for going beyond his authority, in terms of feeling free to disregard this older decision”. Were she less polite, prof. Mathen could have described the Supreme Court as delivering a benchslap to the trial judge, at once gratuitous and telling. Gratuitous, because this part of the Court’s reasons is, in my view, obiter dicta ― it is not part of the reasoning that’s necessary to the decision, which is based on the court’s own re-examination of the constitution and relevant precedent (including, as we’ll see, a departure from Gold Seal). Telling, because the disparagement of history is of a piece with the Court’s broader approach to the constitution, on which more below.

Embarking on its own analysis of s. 121, the Court repeats that a robust reading of this provision would call into question much existing regulation. But, it concludes, such a reading is not required. The constitutional text is “ambiguous, and falls to be interpreted on the basis of the historical, legislative and constitutional contexts”, [54] ― though it is mostly the latter that does the work in the Court’s reasons.

Historical context, in the Court’s view, is inconclusive, because different visions of what form of economic union Confederation would implement were presented by the political actors at the time (none of whom the Court actually quotes). Although it duly notes that “in drafting s. 121, [the framers of the constitution] chose the broad phrase ‘admitted free’ rather than a narrower phrase like ‘free from tariffs'”, [64] the Court insists that “[w]e do not know why they chose this broader, and arguably ambiguous, phrase”, [64] and concludes that “the historical evidence, at best, provides only limited support for the view that ‘admitted free’ in s. 121 was meant as an absolute guarantee of trade free of all barriers”. [67; emphasis in the original]

This is bizarre. Surely we can tell that, if the framers were consciously choosing between a narrower and a broader versions of a constitutional ban on barriers to trade, they chose the broader because the narrower did not capture all the barriers they meant to prohibit. As Benjamin Oliphant and I explain in our article on “Originalist Reasoning in Canadian Constitutional Jurisprudence“, the Supreme Court is no stranger to the “originalist inference” ― reasoning from a choice made during the framing of a constitutional text between competing proposed versions of a provision. The inference seems obvious here, but the Court avoids it. Even more remarkably, the Court also ignores the injunction in Bedford that appellate courts are not to re-assess “social and legislative evidence”, [49] including expert evidence, presented at trial. While the wisdom of this injunction is highly questionable, the Court is, admittedly not for the first time, simply ignoring relevant precedent, without bothering to either distinguish or overrule it.

The “legislative context” that the Court refers to is the placement of s. 121 in a Part of the Constitution Act, 1867 that largely deals with financial issues. The Court considers that  its other provisions “attach to commodities and function by increasing the price of goods”, suggestion that s. 121 does not “capture merely incidental impacts on demand for goods from other provinces”, rather that “direct burdens on the price of commodities”. This might be the Court’s best argument, though it may also be that, as the trial judge found, s. 121 was put where it was simply because this was as good a place as any other in the Constitution Act, 1867. Be that as it may, the Court itself does not seem to attach all that much importance to its conclusion on this point.

The heart of the Court’s reasoning is its discussion of the principle of federalism, which it finds to have two implications of particular relevance to the question of the constitutionality of barriers to inter-provincial trade. One is the exhaustiveness of distribution of powers between Parliament and the provinces. The other is the idea of a balance between the powers of the two levels of government ― and the Court’s role in maintaining that balance. As to the former, the Court insists that there must be no “constitutional hiatuses — circumstances in which no legislature could act”. [72] For any given policy ― including the imposition of barriers to inter-provincial trade ― there must be a level of government competent to enact it, alone or at least in “co-operation” with the other. As to the latter, the Court quotes F.R. Scott for the proposition that “[t]he Canadian constitution cannot be understood if it is approached with some preconceived theory of what federalism is or should be”, [82] and insists that, rather than “a particular vision of the economy that courts must apply”, federalism “posits a framework premised on jurisdictional balance that helps courts identify the range of economic mechanisms that are constitutionally acceptable”. [83]

Here, the Court contradicts both the constitution and itself. Constitutional hiatuses are not anathema to federalism. They exist: in section 96 of the Constitution Act, 1867 (which limits the powers of both Parliament and the legislatures to interfere with the independence and jurisdiction of superior courts); in sections 93(1) and (2) (which limit the provinces’ ability to interfere with minority rights in education, without allowing Parliament to do so); and, even on the Court’s restrictive reading, in s. 121 itself. And then, of course, there is the giant constitutional hiatus usually known as the Canadian Charter of Rights and Freedoms, as well as the smaller but still significant one called section 35 of the Constitution Act, 1982. As for the court’s disclaimer of authority and desire to impose a particular vision of federalism or the economy, it is simply laughable. The idea that federalism requires judicially-imposed “balance” rather than the respect of the letter of the constitution, and any conceivable form of economic regulation must be able to be implemented are precisely the sort of preconceptions that the Court pretends to banish from our constitutional law.

Oblivious to its own incoherence, the Court claims that federal balance would be undermined, and a “constitutional hiatus” created, by an overbroad interpretation of s. 121. Instead of “full economic integration” [85] or “absolute free trade”, the Court propounds what it presents as a compromise:

s. 121 … is best conceived as preventing provinces from passing laws aimed at impeding trade by setting up barriers at boundaries, while allowing them to legislate to achieve goals within their jurisdiction even where such laws may incidentally limit the passage of goods over provincial borders. [91]

The notion of impediment to trade is seemingly a broad one, extending to any provincial law that “imposes an additional cost on goods by virtue of them coming in from outside the province”, [108] or indeed bans inter-provincial importation outright. But, crucially, only laws “aimed at” creating such impediments are prohibited by s. 121, and this will be an extremely narrow category. In effect, it seems that only laws serving primarily “purposes traditionally served by tariffs, such as exploiting the passage of goods across a border solely as a way to collect funds, protecting local industry or punishing another province” will count ― and even that “depending on other factors”. [111] A law having a “rational connection” [113] to some other regulatory purpose, such as “protecting the health and welfare of the people in the province”, [112] or most any other conceivable regulatory objective, will survive. The law at issue survives because it is part of a regulatory scheme intended “to enable public supervision of the production, movement, sale, and use of alcohol within New Brunswick”. [124] Its effects on inter-provincial trade in liquor coming to New Brunswick are merely “incidental”, and constitutionally permissible.

This is wrong in many ways. As a starting point, the Court is answering the wrong question. The issue is not how s. 121 is “best conceived”, but what its purpose is, and how that purpose can be given effect. As Randy Barnett and Even Bernick write in a their essay on purposive constitutional construction (which I reviewed here),

[t]o formulate a rule with reference to the function that the relevant provision is designed to perform is not a matter of making the law “the best it can be” but giving effect to the law as best one can. A judge who decided a case on the basis of some other reason—however normatively appealing that might seem—would be departing from the law entirely. (27)

Second, the Court is wrong to claim that its approach to s. 121 is consistent with precedent. However narrowly it construed s. 121, Gold Seal at least maintained an outright prohibition on inter-provincial tariffs. Following Comeau, tariffs are fine ― provided that they are rationally connected to some regulatory scheme that can be spun to appear to be directed a public health and welfare objective. So much for stare decisis. Most importantly though, as Malcolm Lavoie points out in a CBC op-ed, the Court’s “approach practically nullifies Section 121”, because legislation primarily intended to deal or interfere with inter-provincial trade is already something that provinces cannot enact ― if anyone can, it is Parliament, under section 91(2) of the Constitution Act, 1867. (Professor Lavoie, it is worth noting, is the author of the most important article on the Comeau litigation, which the Court ignored, as it ignored all other scholarship touching on the case, as well as recent work on constitutional interpretation more broadly).

* * *

What causes the Court to re-write the Constitution Act, 1867 (while insisting that it is not making a policy decision), ignore precedent (while admonishing the trial judge for doing so), all in the name of a quest for a federal balance that it is quite different from the one the framers of the constitution struck (while denouncing the imposition of pre-conceived notions of federalism)? Emmett Macfarlane, writing for Maclean’s, denounces Comeau as “craven”, the result of “politicized timidity”. He is not wrong about this (though I think he is in his general denunciation of the federalism jurisprudence), but let me be more specific. In my view there are two (loosely related) problems with the way the Court decided Comeau: its pro-regulatory bias, and approach to constitutional interpretation.

The Court’s bias in favour of regulation appears in the introduction of both the decision as a whole (at [3], quoted above) and that of the substantive part (at [51], in similar terms). The Court is preoccupied by the fact that s. 121 might prevent the enactment of some forms of regulation. It is this, rather than the more general notion of “constitutional hiatuses” that leads it to narrow s. 121 into oblivion. As noted above, hiatuses exist, and the Court is actually quite fond of expanding them, s. 96 and the Charter especially. It is the prospect of constitutional limits on economic regulation that makes the Court suddenly desirous to ensure that Canadian legislatures can make or unmake any law whatever.

As for the Court’s interpretive method, it is implicitly, though not explicitly, living constitutionalist. In an appendix to the “Originalist Reasoning” article, Mr. Oliphant and I wrote that in Comeau the Court “be faced with a stark interpretive choice between a very strong originalist case”, which prevailed at trial, “and arguments based (perhaps paradoxically) both on stare decisis and what may be perceived as the needs, or at least the expectations, of current society”. These perceived needs are reflected in the Court’s pro-regulatory bias which causes it to impose its own vision of federalism. And doing so is all the easier if historical evidence can be treated as less significant and worthy of deference than equivalent social scientific evidence, twisted, or even ignored.

* * *

As I wrote in an essay published last year in Diritto Pubblico Comparato ed Europeo, the well-documented hefty costs of the regulatory schemes which the Supreme Court thought it so important to preserve from constitutional challenge, and the fact that this cost is, in many cases, disproportionately borne by the most economically disadvantage members of Canadian society, ought to remind us that “living constitutionalism can come at a price, not only to abstract ideals such as the Rule of Law, but also to individuals and families, including, and even especially, to the most vulnerable”. (644) To be sure, we can in theory demand that our politicians enact inter-provincial free trade even if our judges will not impose it. But this argument could be made in response to literally any constitutional claim. The raison d’être of an entrenched, judicially enforceable constitution is that the political process sometimes fails to translate just demands, and indeed even popular demands, into legislation, due to either the tyranny of self-centred majorities, or the well-organized resistance of self-interested minorities. Section 121 of the Constitution Act, 1867 was enacted in recognition of this reality. The Supreme Court presumes to update our constitution, but it lacks the wisdom of those who wrote it.

It has been said, perhaps unfairly, that Viscount Haldane was “the wicked stepfather of the Canadian Constitution“. The Supreme Court deserves to be called the Constitution’s spoiled child. This child demands that its parent conform to its demands, and throws tantrums whenever it does not. Unfortunately, too many people find this child’s petulance endearing. Perhaps Comeau will convince them that it must, at long last, be made to behave.


No, constitutional conventions cannot stop free trade within Canada

I didn’t write about the “Free the Beer” decision, R. v. Comeau, 2016 NBPC 3, when it came out this spring. It took me a very long time to read, and others beat me to it ― notably Benjamin Oliphant, to whose excellent analysis over at Policy Options Perspectives there is not much to add. There is one specific point, however, which concerns a pet peeve of mine, and which I do not think others have addressed, which in my mind justifies my doing so here, however belatedly. The point in question is the government’s argument that a constitutional convention meant that section 121 of the Constitution Act, 1867 ought not to be invoked to strike down legislation erecting barriers to inter-provincial trade.

Justice LeBlanc rejected this argument, just as he rejected the rest of the Crown’s submissions on his way to concluding that New Brunswick’s rules prohibiting the importation of alcohol from other provinces are unconstitutional because contrary to section 121. But although there is much to like about his decision overall, I think there is a bit more to say about this particular point than he did.

The Crown claimed, on the basis of an expert report by a political scientist, that the ever-evolving Canadian federalism had developed in such a way that

governments, rather than the courts, have taken on the lion’s share of responsibility for the management of the federation. This is accomplished in part by the courts’ recognition of constitutional conventions and by a judicious deferral to governments to maintain the balance of powers. [153]

One of the conventions in question is, according to the Crown’s expert, the “disuse” [169] of section 121:

governments do not use section 121 to challenge the protectionist policies of other governments. As such perhaps a convention has formed whereby section 121 is effectively rendered inoperative. [171]

The expert, moreover, saw section 121 as a sort of spent transitional provision, arguing that it is rather a convention that prevents the imposition of customs duties at provincial borders.

Justice LeBlanc responded by pointing out, quite rightly, that

[o]nce the Supreme Court of Canada strictly interpreted section 121 [in Gold Seal Ltd. v. Alberta (Attorney-General), (1921), 62 S.C.R. 424, as applying only] to custom duties, there was in reality nowhere else for the section to go. It strictly prohibited custom duties and nothing else. Its disuse became merely a matter of practice or custom. It was not possible for the section to be interpreted in any way to come to the aid of any other governmental policy or strategy.

In other words, Supreme Court precedent limited the scope of section 121 ― though it certainly did not abolish it, so that it is fanciful to claim that a constitutional convention has been doing the work that this provision has always done ― and it is for that reason that it was no longer invoked. That is true, so far as it goes, and it is understandable that a judge would say no more in the course of an opinion that is already quite long enough. But, as I noted above, there is more to say here.

It is worth pointing out that the Crown’s reliance on constitutional conventions in the course of an argument is a pretty remarkable thing. On an orthodox view, constitutional conventions are not enforceable by courts. The Crown analogized section 121 to the provisions of the Constitution Act, 1867 that enable the Governor General (acting on the advice of the federal government, of course) to disallow provincial legislation, which are rendered inoperative by a constitutional convention. Yet the Supreme Court expressed the view, in Reference re The Power of the Governor General in Council to Disallow Provincial Legislation and the Power of Reservation of a Lieutenant-Governor of a Province, [1938] S.C.R. 71, that these provisions were valid an in force as a matter of law. Similarly, in the Patriation Reference the Supreme Court said that conventions were not legal rules. So any attempt to invoke conventions as a sword rather than a shield (arguing that a claim should not be entertained because it asks the court to enforce conventions) faces an uphill battle, and indeed seems pretty desperate. It is telling, I think, that the Crown chose to make such an argument in Comeau.

Now, my own opinion is that the orthodox view that there is a sharp distinction between conventions and law is unfounded. Fabien Gélinas and I have suggested that, at least, conventions should inform the interpretation of the provisions of the written constitution. In a paper called “Towards a Jurisprudence of Constitutional Conventions”, (2011) 11:1 OUCLJ 29, I went further and argued that courts could actually treat conventions the way they treat common law constitutional rules, subject to justiciability concerns. While it is far from obvious that courts would endorse either of these approaches, and not very clear that the Crown in Comeau made much of an argument to show that they ought to, let’s assume that the court would have been bound to take a relevant convention into account in one way or another. Of course the question is whether there is such a convention here. Justice LeBlanc’s reasons suggest that the answer is “no.” But they ― understandably ― do not go into any detail on this point. A closer look shows that the Crown’s argument is completely off-base.

All constitutional conventions limit or eliminate he discretion that political actors seem to enjoy pursuant to other constitutional rules. For example, the law of the constitution leaves the sovereign with the choice to assent or not to bills that have passed the House of Commons and the Senate, but convention eliminates this discretion. The sovereign must assent. Geography does not figure among the criteria which the Supreme Court Act provides for the appointment of Supreme Court judges, other than those from Québec, but convention reduces the government’s discretion as to the advice it gives the Governor General by supplying additional geographic requirements.

What about the alleged convention here? The Crown’s expert points out that governments have refrained from suing each other on the basis of section 121. But even if that forbearance could be said to have acquired the status of a conventional rule, this convention could apply to governments ― the political actors whose behaviour contributed to the alleged rule’s emergence ― and only to governments. Not to citizens. To repeat, conventions stipulate how political actors exercise discretion. They do not dictate the behaviour of citizens. So while a convention may in effect nullify constitutional provisions that only empowers a political actor, such as those dealing with the disallowance power, they cannot “render[] inoperative” provisions that confer rights on citizens.

The Crown’s argument assumes, without even attempting to demonstrate, that section 121 is a provision that only concerns governments. But the assumption is unwarranted, and indeed galling. Constitutional provisions limiting the power of governments, such as section 121, exist in order to preserve the liberty of the citizens. In Attorney General of Nova Scotia v. Attorney General of Canada, [1951] S.C.R. 31, Chief Justice Rinfret wrote that even if Parliament and the legislatures agree to modify the constitutional division of powers by resorting to delegation, they cannot do so, because

[t]he constitution of Canada does not belong either to Parliament, or to the Legislatures; it belongs to the country and it is there that the citizens of the country will find the protection of the rights to which they are entitled. (34)

As the Chief Justice pointed out,

[i]t is part of that protection that Parliament can legislate only on the subject matters referred to it by section 91 and that each Province can legislate exclusively on the subject matters referred to it by section 92. (34)

But another part of that protection, of course, is that when the constitution removes a legislative power from both Parliament and the provinces, neither can arrogate such a power to itself, even with the connivance of the other. This is true of the power of constitutional amendment, for instance, and of the violation of Charter rights. And it is equally true of section 121. Were a court to accept to Crown’s (un)conventional argument to the contrary, it would transform the Canadian constitution from protection of the citizens’ freedom into a plaything for governments intent on limiting that freedom.