Corps de l’article


This article considers the content of the unwritten principle of democracy and its potential relevance in Canadian constitutional interpretation. The unwritten constitutional principles of federalism, constitutionalism and the rule of law, democracy, and the protection of minorities enunciated into being by the Supreme Court of Canada (the Court) in the Secession Reference[1] have received extensive academic attention. Much yet remains unknown, however, about the democracy principle. Post-Secession Reference, important and substantial scholarly work has taken on justifying or critiquing the use of unwritten principles in constitutional interpretation. The courts have assessed the meaning of the principles of federalism, the rule of law, and judicial independence in some detail. The content of the democracy principle has received less elaboration, apart from the Court’s comments in the Secession Reference, and indirect definition implied from judicial interpretation of the related principle of parliamentary sovereignty.[2]

Democracy is an “essentially contested concept,” meaning one that “inevitably involve[s] endless disputes about [its] proper uses on the part of [its] users.”[3] There are radically competing notions of what “democracy” entails. Günter Frankenburg writes that “[t]he conceptual histories of democracy span more than 2,500 years and refer to a variety of normative orders, institutional arrangements of political decision-making, social and economic structures, and basic values of a community.”[4] The contested nature of the concept and its political salience muddy what vision of “democracy” should prevail in constitutional interpretation.[5] Perhaps resulting from this conceptual uncertainty, the scholarship has rarely tackled head-on the meaning of democracy as a legal principle in Canada.[6] Delineating the content of the unwritten principle for the purposes of constitutional interpretation from the concept of democracy in the abstract is, therefore, an important task.

Focusing on the content of the principle of democracy is also particularly relevant at this moment in Canadian constitutionalism. First, there is a lack of clarity regarding its content in the case law. Claimants have raised the principle recently in constitutional litigation in a number of different contexts, from electoral reform and referendums[7] to municipal elections and electoral boundaries,[8] among others. The unwritten principle of democracy remains of direct contemporary relevance.

Second, the Canadian Constitution is showing its age. The Constitution Act, 1867 and even the more recent Constitution Act, 1982 simply do not address important aspects of contemporary democracy, no matter the interpretive approach adopted to deciphering the text. There are gaps in the constitutional text in relation to the conduct of elections and representation, for example.[9] While textual silence in a constitution can of course be deliberate,[10] the basic fact of representative democracy remains underspecified in the Canadian Constitution. The lack of specificity with regard to “democracy” in the text means the unwritten principle has a lot of work to do. Expectations for constitutional protection of democracy have shifted quite rapidly since 1982, as have global trends in constitutional design of democracy. Many influential constitutions, for example, protect a much larger set of political rights and electoral institutions than the text of the Canadian Constitution.[11]

Third, hovering in the background of most debates in Canadian constitutional law is the reality that constitutional amendments requiring any degree of federal-provincial consensus are nearly impossible, politically.[12] Put bluntly, we are stuck with the text that we have for the near future. Without a realistic chance of amendment, there will continue to be pressure on courts to resolve constitutional disputes where the text provides little clear guidance or no longer reflects the social, economic, or demographic facts of Canada.[13] Given that the text in relation to democracy has gaps and is out of date in relation to global practice, there is likely to be ongoing impetus to resort to the unwritten principle.[14]

This article argues that we should interpret the unwritten principle as embodying a “thin” or procedural account of democracy tied to meaningful participation, rather than a “thick” version imposing specific outcomes or broader obligations. The article proceeds as follows. Part I details the ways in which “democracy” has manifested itself in the Canadian Constitution and the gaps in the text. Part II presents the most forceful critiques raised about the use of unwritten principles in constitutional interpretation, namely, that (1) judicial review based on unwritten principles lacks legitimacy, and (2) there is an “interrelatedness” problem, in that the principles are incoherent in relation to one another.[15] Part III engages with these critiques to argue that a “thin” or procedural version of democracy best responds to the legitimacy and incoherence critiques. In relation to the legitimacy critique, a thin version of democracy is compatible with the existing constitutional text and has less scope for the arbitrary imposition of any particular judge’s policy preferences. In relation to the incoherence critique, I argue that a thin version of democracy is a better fit with the other principles. A thin version of democracy would require meaningful participation through fair procedures in what I call democratic events. This interpretation of the principle furthers the values of political accountability and political equality. Part IV then applies this thin understanding of the unwritten principle in two contexts either not addressed, or only partially so, by the constitutional text: 1) referendums; and 2) municipal elections. I outline how the unwritten principle should result in fair procedures in both of these democratic events. I conclude the article by briefly considering the future of the unwritten principle of democracy.

I. Democracy in the Canadian Constitution: A Brief Overview

The Constitution Act, 1867 is parsimonious in its treatment of democracy. It does not expressly determine democracy’s contours in any great detail, largely because of the assumption that Canada would follow the British model of an unwritten constitution. Accordingly, parliamentary sovereignty was the pre-eminent feature of democracy.[16] The text creates a bicameral Parliament with the qualifications for Senators[17] set out along with other features of the Upper House.[18] Constitutional conventions largely determine the relationship between the different branches of government.[19] Parliament controls its own internal operations through parliamentary privilege[20] and the operation of the conventions. The provisions on federalism delineate separate, democratically elected orders of government accountable to different publics.[21] Where the text does go into detail is on representation, particularly geographic representation. The text establishes the names, number, and boundaries of the early electoral districts, as well as the number of seats assigned to each province in the House[22] and Senate.[23]

The Constitution Act, 1982 addressed the functioning of legislatures by setting limits on their traditional freedoms.[24] Parliament and each legislature are required to sit at a minimum once per year[25] and cannot continue more than five years between elections,[26] except in circumstances of “war, invasion or insurrection.”[27] The procedures on constitutional amendment in Part V of the Constitution Act, 1982 also reinforce the focus from 1867 on representation of the provinces in a diverse federation. The smaller provinces have a guarantee in Part V that the existing rule that they can have no fewer MPs than they have Senators can only be changed with unanimous consent.[28] The “proportionate representation” of the provinces in the House set out in section 51 of the Constitution Act, 1867 can only be altered according to the 7/50 formula.[29] Part V as interpreted by the Court also locks in the method of selecting Senators, as the Court has read down the unilateral amendment procedure in section 44.[30]

The Canadian Charter of Rights and Freedoms brought with it a host of entrenched rights and freedoms of direct relevance to elections.[31] Section 3 provides the right to vote in federal and provincial elections as well as to stand as a candidate. Parliament and the provincial legislatures frequently treated the right to vote as a privilege within their purview to restrict or rescind, prior to 1982.[32] Outright bans on voting by a defined group of citizens appear impermissible.[33] The only remaining group of citizens barred from voting are individuals under the age of eighteen.[34] Courts have also interpreted the provision as protecting much more than simply the right to cast a ballot. Section 3 protects “effective representation” in the allocation of the number of electoral districts and the design of their boundaries.[35] The “meaningful participation” of electors through the political parties that they choose to support is also included.[36] The text appears on its face to limit the application of section 3 to federal and provincial elections. It has been interpreted to date to exclude territorial, municipal, school board, and band council elections.

Freedom of political expression also has obvious and direct implications for democracy. The courts have repeatedly held that political expression, including the freedom to criticize the government, is at the very core of democracy. However, the Court has upheld spending caps on political advertising as reasonable limits on expression.[37] Where section 3 does not apply, as with a school board election,[38] referendum,[39] or municipal vote,[40] claimants have at times argued that the content otherwise present in the right to vote should also be found within section 2(b), though largely without success.

The right to equality in section 15 also indirectly relates to elections. Discrimination against voters on prohibited grounds is barred. Section 15 has played an increasing role in litigation surrounding band council elections[41] under the Indian Act,[42] where section 3 does not apply.

The text of the Constitution with regard to democracy reflects the times at which the particular provisions were drafted and its British colonial heritage. There is much more that could be said about the operation of Canadian democracy and the particular approach reflected in the text of the Constitution. For my purposes, however, it is sufficient to draw the following conclusions.

First, the Canadian Constitution is incomplete with regard to democracy. The text still contains gaps.[43] Not all matters ignored or only covered partially by the text have been resolved through constitutional conventions or the broad language in sections 3, 2(b), and 15. The most obvious gap is that the text does not directly specify the electoral system. Despite the size of some cities and the current importance of municipal government, much of the structure of municipal elections remains vulnerable to the whims of provincial political majorities.[44] The inheritance of the Westminster representative model de-emphasized mechanisms of direct democracy, such as referendums. A host of issues covered in many newer constitutions, including the existence and powers of a non-partisan commission to administer elections, voter registration, electoral district design, and so on, are simply not included.[45] Newer constitutions are much more likely to contain detailed prescriptions in the text. From a comparative perspective, the focus in contemporary constitutionalism on ensuring fair election administration is absent.[46]

Second, the text of the Charter in relation to democracy and elections requires much from courts. Courts have had to go beneath the text to the foundational values of the Constitution and engage in structural reasoning in order to reach outcomes that protect against the imposition of unfair electoral rules by legislative bodies. Admittedly, the jurisprudence under section 3 striking down bans on voting by adult citizens, most recently in Frank, do largely turn on the text.[47] Section 3 enfranchises “[e]very citizen.” In Frank, the Court took this phrasing to mean the framers were signalling that restrictions on voting based on residence violated the right to vote and must be justified under section 1. It is hard to say credibly, however, that the courts have stuck closely to the text in the jurisprudence on section 3 and section 2(b). The Court has made it clear that the right to vote, for example, means much more than simply the right to put a piece of paper in a ballot box, and instead protects the background factors that ensure an election is fair.[48] It is necessary for courts to continually protect the background structures of democracy so that the public can hold elected representatives accountable at election time.[49] A consequence of this conclusion, however, is that courts are obliged to engage in structural reasoning that requires consideration of a host of factors—including what kind of democracy should be protected in Canada—that are not dictated squarely by the text.

Third, no one particular theory of democracy has prevailed. This conclusion is perhaps not surprising given the multiple constitutional documents, disparate provisions, and the evolution of the case law over a relatively long period. The Constitution contains egalitarian measures, such as the right to vote provided in relatively untrammelled terms to all citizens, and those that are elitist, such as the unelected Senate. It creates multiple democratic systems within the federation, which need not operate in identical fashion. The Charter does set clear markers for democratic rights and freedoms, but there remains room for experimentation and different institutional approaches, within the framework of representative democracy and parliamentary supremacy. There is no one uniform vision for democracy. The Court’s definition of the unwritten principle of democracy in the Secession Reference reflects these plural heritages. The Court’s summary of the principle contains elements that focus on its Westminster lineage and representative democracy, others that accept the particular idiosyncrasies of the Canadian experience such as the unelected Senate, and gestures toward liberal, direct, and deliberative democratic theories.[50] The Court sought to avoid boxing itself in to any particular account of democracy.

All of this suggests that there is ample space for the unwritten principle of democracy to operate in the Canadian constitutional order. The presence of gaps, broad language in the text, and the absence of any single prevailing theory, in an aging Constitution that cannot easily be amended, all indicate an important functional role for the principle in constitutional interpretation.

II. Critiques of the Unwritten Principles

If there is space for the principle of democracy to do work in constitutional interpretation, the question that then arises is what should be its content? I will argue in Part III that a “thin” account focused on meaningful participation and procedural fairness is the most appropriate—rather than a “thicker” theory—given that the unwritten principle will be applied by judges engaged in constitutional interpretation. Before arriving at that argument, it is necessary to consider the critiques in the scholarship on the use of unwritten principles in constitutional interpretation, which I do in Part II below. In my view, a thin, procedural account of the principle of democracy is better able to respond to the arguments against the use of unwritten principles more generally.

We now have more than two decades of scholarship post-Secession Reference that has engaged with the Supreme Court’s interpretation of the role of unwritten principles in the Canadian constitutional order.[51] The two major lines of critique of the use of unwritten constitutional principles remain those set out by Jean Leclair in 2002, which he termed the legitimacy and interrelatedness arguments.[52] In brief, in the first argument, it is illegitimate for courts to apply unwritten principles in constitutional interpretation or, more narrowly, to apply them to invalidate statutes duly passed by the legislature. In the second set of arguments, the uncertain relationship and shifting boundary lines between principles, given their unwritten nature, inevitably results in incoherence and arbitrariness in their application. I engage with each of the two critiques in turn.

A. Legitimacy Critique

The legitimacy critique has multiple strands. They focus on the unwritten nature of the principles, separation of powers concerns, and the worry that they permit natural law to impede the democratically legitimate positive legal order. I set each out in turn.

One part of the legitimacy critique disapproves of the unwritten nature of the principles.[53] At its root is the claim that unwritten principles are judge-made constitutional law, fundamentally at odds with the notion of a written constitution. Unwritten principles are said to diverge from constitutional conventions, which have always been part of the Canadian constitutional order, because the latter are only politically enforceable. Like constitutional conventions, in this argument, the boundaries of principles are inevitably fuzzy, given their unwritten form. The legal enforcement of unwritten principles through judicial review, however, makes them problematic in a way that conventions are not.

Another strand of the legitimacy critique emphasizes separation of powers concerns. As this argument goes, unwritten principles transfer power to the courts that is properly within the ambit, first, of the framers, and then, of legislators. They provide the judicial hook with which to pull aside text that the courts find inconvenient. The acceptance of unwritten principles in the constitutional order, especially if used to invalidate statutes, allows judges to impose their own policy preferences, rather than those of the electorate or their representatives.[54] Former Chief Justice Beverley McLachlin articulates the heart of the claim as being that unwritten principles are a “barely concealed power grab by activist judges.”[55] The unwritten principles provide nearly infinite room from which to stray from the constitutional text in order to suit judicial whim, on this account.

A third strand of the legitimacy critique is what Mark Walters has called the “naturalist” argument.[56] In this argument, unwritten principles allow judges to smuggle natural law into a positive legal order that is otherwise antithetical to natural law. Unwritten principles are often seen as a “modern reincarnation of the ancient doctrines of natural law.”[57]

All three of these critiques rest on the same fundamental point—that constitutional interpretation relying on unwritten principles is illegitimate and should be avoided or, at least, constrained. They also share the common view that there is a hierarchy of misuses of unwritten principles. The most egregious misuse for those concerned about legitimacy is when courts apply unwritten principles to invalidate the clear will of the legislature as expressed in a statute. Their use to interpret existing text or to fill in true “gaps” in the text is problematic, on this view, but generally less so than invalidating statutory text.

The responses to legitimacy-based critiques of unwritten principles have been legion, from a variety of approaches.[58] Martin Loughlin argues that constitutional silence is inevitable, and even desirable, which implies that unwritten principles are as well.[59] T. R. S. Allan claims that “[t]he written law is always underpinned and informed by the unwritten, reflecting the discourse of reason.”[60] From a comparative perspective, Vivek Krishnamurthy writes that unwritten principles operate “[i]n every modern democratic constitutional order,”[61] and that the Canadian principles are relatively uncontroversial, as “most any constitutional court in a modern federal democratic state would probably have come up with the same list.”[62] David Dyzenhaus states that in order to “uphold the rule of law ... [judges must] interpret the positive law of a legal order in light of their understanding of unwritten constitutional values.”[63]

Those scholars who argue that unwritten principles are inevitable even in the presence of written text have the better argument over those who claim that their use is illegitimate. Even if there are no true “silences” in the positive law, judicial reasoning about the content of the text on the page must rely to some extent on background values. The lack of positive law with a “determinate content”[64] is a circumstance that confronts Canadian judges in particular. Canadian positive law is inadequate to resolve many constitutional disputes, given the gaps in text, including around what content to attribute to the democracy-related provisions in the Constitution.

B. Interrelatedness

A second major line of critique is what Jean Leclair calls “interrelatedness.”[65] On this view, the result of having multiple, unwritten principles is incoherence. The inherently abstract nature of each principle and the unclear boundary lines between them are a recipe for conceptual confusion.

Leclair argues that “some of these [unwritten] principles are so abstract, that, by themselves, they provide no clear answer. Democracy and federalism, for instance, can be understood in many ways.”[66] On this view, if the principles are to have legal force,[67] and cannot be exclusively understood in terms of ideal-type political theory,[68] then they must have a discernible content and be justiciable in the sense of being subject to consistent judicial application. Being principles, however, they must be general in character. Their inevitable abstractness means they are likely to slip through the hands of anyone attempting to fix a particular, concrete meaning. Potential disputes about the boundaries of the four unwritten principles from the Secession Reference—democracy, federalism, the rule of law and constitutionalism, and the protection of minorities—are inevitable. Where does democracy begin and federalism end, for example? On this account, the inevitability of conceptual dissonance undermines the very presence of unwritten principles. The risk of conflict between principles means that they may even be “irreconcilable.”[69] The fact that there are more principles than the four set out in the Secession References magnifies the possibilities for conflict.[70] Their abstractness and possible irreconcilability militate against their usefulness as legal principles.[71]

This irreconcilability raises another problem, which is that courts may favour one principle over the others in the case of conflict. John Tasioulas has suggested that as a response to conflict between unwritten principles, everything deemed desirable is sometimes crammed into one principle at the expense of the others. One of the unwritten principles may eventually dominate and, indeed, come to incorporate all of the perceived beneficial content of other related principles.[72] Tasioulas has thick versions of the rule of law in mind here, but the risk that some principles will atrophy while others predominate has more general resonance. Jo Murkens appears to advocate for democracy to take on the role as the pre-eminent principle in the United Kingdom, for example, putting it ahead of its competitors, especially parliamentary sovereignty.[73]

The Court in the Secession Reference anticipated such critiques about the relationship of the principles to one another. It held that “[t]hese defining principles function in symbiosis. No single principle can be defined in isolation from the others, nor does any one principle trump or exclude the operation of any other.”[74] The Court emphasized the mutually reinforcing content of the principles to head off attacks about fuzzy boundaries and interrelatedness. It rejected the idea of a “trump” in light of concerns of the type raised later on by Tasioulas. The Court did live up to its stated method in the way that it defined the principles, for example by holding that majority rule is a key part of democracy, but that democracy was not equivalent to majority rule, so as to protect territorial or other minorities.[75]

The problem with this schema for the unwritten principles is that it multiplied concerns about legitimacy and incoherence. Apart from respect for federalism, the general notion of citizen participation and a universal franchise, and protection of minorities from tyranny of the majority, the Court drew few firm limits on the principle of democracy. The precise relationship between democracy and parliamentary sovereignty, for example, remains unclear. In the interrelatedness critique, the Court provided a recipe for conflict between the principles, rather than for avoiding one.

C. Implications for the Content of Unwritten Principles

The critiques and defences of unwritten principles that emerged post-Secession Reference focused largely on whether the courts should use them in constitutional interpretation and, if so, how they should interact with the positive legal order. These are weighty matters deserving of the sustained scholarly attention that they received. There is also in my view a need to focus on the particular content of the unwritten principles, at least with regard to democracy.

There are two main ways in which the unwritten principle of democracy can feature in constitutional interpretation. It can inform the interpretation of the written constitution, for example the provisions of the Charter related to voting and political expression. It is difficult to envision persuasive reasoning by courts about the meaning of the right to vote, freedom of political expression, or equality in political life without some deeper sense of democracy animating the interpretation of the text. More controversially, the principle could also have its own independent normative force. This second use of the principle is particularly relevant if there are constitutional gaps that need filling because of the evolution of subject matter covered in the text beyond anything contemplated by the framers or the emergence of truly new issues.

If we accept the general point that unwritten principles are inevitably part of constitutional interpretation, because no positive legal order can exist without reasoning behind the text and all such orders are inevitably incomplete, determining the content of such principles becomes an important project.[76] While I agree with the general point that unwritten principles are necessary and unavoidable features of constitutional interpretation, in my view we must still respond to the legitimacy and interrelatedness critique in how we define the content of those principles. The two main critiques have purchase if principles mean whatever judges want them to mean in the moment, or if their scope is unbounded. In other words, the content of the unwritten principles must be defined, but also constrained. Constraining the content of the democracy principle, so that it may fulfill its purpose in constitutional interpretation without lending credence to the legitimacy and interrelatedness critiques, remains an unfinished task.[77] Courts and scholars have debated the possible meanings of federalism[78] and the rule of law[79] as legal principles in the Canadian context quite extensively. By contrast, the Secession Reference provided only a vague outline as to the meaning of democracy in constitutional interpretation and advances since then have been incremental.

The challenge is to define the democracy principle so that it has the scope to inform constitutional interpretation in a meaningful way, but narrowly enough so as to not push aside the text or swallow up the other unwritten principles. I turn in the next section to the task of setting out an account of the content of the principle of democracy capable of meeting the legitimacy challenge and the interrelatedness critique.

III. The Content of the Democracy Principle

In this section, I argue for a “thin,” procedural understanding of democracy as a means for limiting the scope of the unwritten principle while still allowing it to do meaningful work in constitutional interpretation. I will first briefly set out the differences between “thin” and “thick” accounts of democracy and argue for why a procedural account is preferable in constitutional interpretation. I will then set out a version of a “thin” account focused on meaningful participation and fair procedures.

A. Thin Versus Thick Democracy

There is a longstanding debate in democratic theory about whether “thick” or “thin” versions of democracy are preferable. In brief, thin versions of democracy emphasize the procedural element of democracy and the institutions that reflect fundamental values, particularly the political equality of adult citizens. This branch of democratic theory puts a premium on the presence of competitive elections as a necessary, though admittedly not sufficient, condition for democracy. Robert Dahl in this tradition writes about the criteria[80] or minimum conditions for democracy to be possible.[81] Phillippe Schmitter and Terry Karl have also famously adopted a thin or procedural account of democracy, stating that “[m]odern political democracy is a system of governance in which rulers are held accountable for their actions in the public realm by citizens, acting indirectly through the competition and cooperation of their elected representatives.”[82] These procedural accounts owe much to Schumpeter, who emphasized the necessity of competition for elected office.[83] Constitutional scholars have resuscitated procedural accounts of democracy in order to justify judicial review that protects the background or “structural” conditions of politics.[84]

Advocates of “thicker” versions of democracy criticize procedural accounts for focusing on inputs rather than outputs. Thin democracy is unduly focused on elections[85] and incorrigibly elite-driven, according to thick accounts. Ronald Dworkin in an exchange with Jürgen Habermas famously dismissed procedural accounts as mere “statistical democracy.”[86] Thick accounts vary in their prescriptions or understanding of what true democracy is, but unite in seeking something more than formal mechanisms for participation, equality enshrined in formal liberal rights and institutions, and competitive elections.

Some thick versions of democracy pick up the procedural emphasis on equality to argue that democracy must entail relatively equal outcomes.[87] W. J. Waluchow argues for a Dworkinian “constitutional conception of democracy” to prevail in understanding the relationship between judicial review and democracy in Canada.[88] Participatory democrats argue that true democracy mandates dramatically increased individual participation,[89] including mechanisms of direct democracy.[90] Feminist theorists have detailed the inadequacies of democracy without deeper commitments to institutional transformation and equality of outcomes.[91]

Among the most prominent counterpoints to thin accounts in recent years has been deliberative democratic theory.[92] While there are many versions, deliberative democrats emphasize reasoned debate and consensus decision-making based on mutually acceptable public reasons, instead of majority-rule procedures such as elections.[93] Rather than aggregating the raw preferences of individuals through voting where the winner takes all power, deliberative democrats seek to implement reasoned, public-minded debate as a precondition for meaningful political decisions based on assessments of the collective good. The Supreme Court’s decision in the Secession Reference reflects this scholarship in its emphasis on democracy as entailing “consent”[94] developed through a “continuous process of discussion,”[95] and by recognizing that “the need to build majorities necessitates compromise, negotiation, and deliberation.”[96]

Thin accounts of democracy highlight the minimum conditions for democracy to exist. This focus provides a set of procedures and institutions whose sufficiency advocates of thick democracy dispute. Proponents of thick accounts, however, typically do not propose to do away entirely with the procedures and institutions set out in procedural versions.[97] They instead believe that true democracy requires supplementing those conditions. There is some common ground then, between thin and thick accounts. Deliberative democrats, for example, generally do not propose to do away entirely with voting in competitive elections as a decision-mechanism, despite their antipathy to preference aggregation.[98]

Thin versions of democracy focused on fair procedures are more appropriate for forming the content of a legal principle to be applied in constitutional interpretation. Whatever the appeal of outcome-focused definitions writ large, the imposition of thick democracy by the courts through the vehicle of an unwritten principle would validate the legitimacy and interrelatedness concerns. The thicker the principle, the more likely there is to be serious disagreement about its content. A procedural account rests on ground that is likely to be more common across the range of political beliefs and theories of adjudication. All democrats do (or perhaps should) endorse the necessity of fair procedures for aggregating preferences as a minimum condition for democracy to exist.[99] There may be disputes on the margins of what constitutes a “fair” election or referendum, but there is likely to be consensus on at least the core of the matter. A narrower account is less likely to face legitimacy problems, as it provides less leeway for courts to impose their own policy preferences or to engage in freestanding reasoning untethered from the actual specifics of the positive legal order.[100] The political branches may possess the accepted authority to implement a thicker version of democracy, but courts do not. A thin, procedural account of democracy is more likely to lead to discernible and justiciable content for the unwritten principle. It is also more likely to be compatible with the existence of other meaningful principles. In short, a thick account has a harder time rebutting the charges levelled against unwritten principles by their scholarly critics.[101]

B. Thin Democracy: Accountability and Equality

We should envision the principle of democracy understood in the way sketched out in this argument as a means of furthering meaningful participation through fair procedures. In Figueroa, the Supreme Court of Canada advanced an interpretation of the right to vote in section 3 of the Charter as requiring “meaningful participation.”[102] One can transport the concept of meaningful participation out of the context of section 3 in order to fill in the content of a procedural account of the unwritten principle of democracy.

Meaningful participation is a useful standard for providing equal opportunities to involve oneself in electoral politics. It does not require particular outcomes, but instead emphasizes the right of citizens to engage fully in the process by requiring fair procedures. In Figueroa, meaningful participation meant the right to partake in federal politics through the vehicle of a small political party, even though that party was not likely to form government.[103] There was no right to have one’s preferences win out in the election, but instead to have an equal opportunity to attempt to sway one’s fellow citizens on policy, to compete for votes, and to have a level playing field upon which one’s preferred political party could operate.

Obviously, no procedural account can entirely sidestep substance or normative considerations.[104] The values furthered by a standard of meaningful participation are, in my view, political equality and accountability. Political equality is furthered by fair procedures tied to meaningful participation because engagement on equal, formal terms sends a message that each participant is worthy of concern and respect. All voters or citizens are political equals in this sense, even if there are differences in their interest in politics or their influence. The Supreme Court recently recognized that denials of the franchise communicate that those targeted are somehow less worthy than those permitted to vote.[105] This insight is an important one. There is expressive content to political rights. They are a signal of full membership in society, rather than partial or semi-citizenship.[106]

An account focused on fair procedures also furthers political accountability. Where the constitutional text does not dictate how relevant aspects of political participation are to be structured, the unwritten principle can be a check on the imposition of unfair procedures. Much of the literature on the Canadian law of democracy focuses on the risk of partisan self-dealing, whereby incumbents or political parties write laws that make their preferred electoral outcome more likely.[107] The text of the Charter should prevent manifestly unfair procedures in federal and provincial elections, such as gerrymandering, or banning participation by rival political parties,[108] or limiting access to the ballot box for citizens.[109] Fair procedures facilitate the ability of citizens to hold their representatives, as well as political parties, to account.

If one accepts the use of a procedural account of the unwritten principle, we must still ascertain to which political practices it may apply. In my argument, the unwritten principle should be in play wherever there is a democratic event. The term democratic event should include elections for representatives in public bodies, but also consultative procedures of direct democracy. Such an approach would mean that the unwritten principle of democracy would apply to municipal, territorial, and other elections for public office,[110] which section 3 does not cover. It would also apply to consultative mechanisms such as referendums, which the text incompletely covers. If an election is held, it should be based on fair procedures. There may be variation between what is required for a federal and a school board election, but the animating spirit of democracy in the Canadian constitutional order should be present for any democratic event.

Even on this vision of a robust role for the unwritten principle, the text would remain paramount. As a general matter, the unwritten principles should not crowd out the text, though they might remain an important source to draw upon in interpreting it.[111] In drawing constitutional limits on legislative action, the text is the preferable option. The principle of democracy also should not cannibalize other related principles, in line with the interrelatedness critique. Parliamentary sovereignty in particular is a limit on the principle of democracy. Courts could not legitimately use the principle against legislative inaction or true, deliberate silence,[112] out of respect for parliamentary sovereignty and the separation of powers.

In order to flesh out the implications of my argument, I consider two case studies in depth here: referendums and municipal elections. While responsive to concerns about legitimacy, requiring fair procedures and meaningful participation on a thin account would still lead to concrete changes that better align with democratic values. The uncertain application of the unwritten principle of democracy to referendums has been a long-standing question, especially post-Secession Reference. The recent controversy in Ontario—over the Legislative Assembly’s change to municipal ward boundaries for the City of Toronto mid-election[113]—directly raised the use of the unwritten principle for municipal elections.

IV. The Scope of the Democracy Principle

A. Referendums

Referendums are a classic instance of direct democracy, traditionally seen as being at odds with representative democracy.[114] Based as it is on the British model, Canadian parliamentary democracy has generally been loath to implement regular referendums. There are no legally enforceable requirements to hold referendums in Canadian constitutional law. The amending procedures in Part V of the Constitution Act, 1982 famously do not require a referendum for constitutional change.[115] The right to vote in section 3 of the Charter does not apply to referendums.[116] Respecting parliamentary sovereignty means that the decision to call a referendum rests with the legislatures, not with courts wielding the unwritten principle.

Referendums have emerged, however, as an important tool of political consultation. There have been only three federal referendums since 1867, with the 1992 vote on the constitutional amendments proposed in the Charlottetown Accord as the most prominent recent example.[117] The provinces have used referendums more frequently, with multiple votes on the secession of Quebec as the obvious example. There has been an emerging pattern in recent years of a political imperative to hold a referendum where electoral reform is on the table. British Columbia has had three referendums on electoral reform over the last two decades,[118] Prince Edward Island has had two,[119] and Ontario one.[120] Whether a referendum is a constitutional convention where major constitutional change or electoral reform are proposed is now a matter of some debate.[121] Globally, democracies hold referendums largely to consult the people when they are contemplating constitutional or quasi-constitutional change.[122]

Surveying their role around the globe, Laurence Morel concludes that there has been increasing “juridicization” of the referendum.[123] Canada is a partial exception to this trend. The courts have interpreted section 2(b) as imposing limits on their conduct once the legislative body has made the decision to hold one.[124] Section 15 applies as well, in theory. The application of the particular provisions of the Charter to referendums, however, has failed to result in judicial oversight of fair procedures or truly meaningful participation. With regard to fair referendums, the central issue is the right to cast a ballot, given that the right to vote in section 3 of the Charter does not apply. The unwritten principle has work to do in ensuring equality and accountability through the vehicles of sections 2(b) and 15.

1. The Right to Vote and Voter Qualifications

The unwritten principle of democracy should also protect the right to vote in a referendum. Prior to the Secession Reference’s invocation of the democracy principle, the Supreme Court ruled bluntly in Haig v. Canada that “there is ... no constitutionally entrenched right to vote in a referendum.”[125] Contrary to Haig, Colin Feasby argues for the extension of section 3 beyond federal and provincial elections.[126]

Canada has a long and desultory history of restrictions on the right to vote. The Charter has swept away most limits on exercising the franchise, with Frank as the most recent example. The fondness of politicians for manipulating the contours of the electorate is a consistent trend in Canadian politics.[127] Outright bans on voting appear too blatant and the courts now deal with them relatively easily, as Sauvé and Frank indicate. The risk of manipulations to the franchise, however, whether in general elections or referendums remains an ongoing one. Restrictions on voting, including in referendums, is a feature of democratic decline in various countries.[128] The most popular tactic is now voter suppression, such as the introduction of strict voter identification rules, differential procedures for voting that favour one group, limitations on early voting introduced in ways that decrease the turnout of specific groups, and so on.

We should not countenance outright bans on specific groups of citizens voting in referendums. A constitutional requirement for an accessible franchise in referendums, as exists for federal and provincial elections, should not be controversial. Voting has expressive content,[129] whether in an election or referendum, and so we should interpret section 2(b) to protect it. It is arbitrary that the courts have not found section 2(b) to protect voting outside of federal and provincial elections. Bans on specific groups voting in a referendum could violate section 15 as well. Whatever their potential application, the doctrine under each of these provisions of the Charter has evolved in ways that make them generally ill suited to ascertain the often-subtle ways in which voter suppression occurs.

Meaningful participation in referendums must require at a minimum a right to cast a ballot. It should also require fair procedures of election administration. If a government chooses to conduct a referendum, then the unwritten principle of democracy should guarantee that it occurs on fair terms and provides opportunities to engage meaningfully in the process. The “gaps” in the Charter’s treatment of referendums have become untenable, given how often they are now used, the importance of the questions they now address, and the risks of interference in the process by self-interested partisans. Referendums are not general elections and serve different functions, but are the next closest thing in terms of importance in the political system. In that context, the constitutional order should require meaningful participation and procedural fairness.

B. Municipal Elections

The unwritten principle of democracy should be seen as playing an important role in constitutional protection of fair municipal elections, as long as it is appropriately constrained in relation to the parliamentary sovereignty of provincial legislatures. The current positive law with regard to municipal elections is deeply unsatisfying if the Constitution is supposed to protect procedural fairness and meaningful participation. The recent controversy surrounding the Ontario government’s decision to decrease the number of wards in Toronto in the middle of an ongoing city election campaign highlighted the imperfect protection offered by the Constitution.[130] While Toronto (City of) v. Ontario (AG)[131] brought much public attention to the issue, in reality the unsatisfying constitutional parameters around municipalities and municipal elections has been a long-standing issue. The unwritten principle can and should fill in the gaps in the existing constitutional text to ensure fair municipal elections. It should also be the principled foundation upon which the doctrines under the applicable provisions of the Charter can evolve.

The constitutional status of municipalities is clear.[132] The Constitution Act, 1867 did not imbue municipal governments with any special constitutional status. They are creatures of the provinces by virtue of section 92(8).[133] The courts have had to confront the question of whether municipalities have evolved to the point where their constitutional importance as democratically elected governments providing important services should be recognized by the Constitution.[134] For example, the case law on the application of the Charter under section 32 to municipalities has left open the possibility that they are the functional equivalents of the federal and provincial governments.[135] While formally administrative entities, municipalities possess a democratic nature that separates them from other such bodies.[136] The special status of municipalities also arises in the redistribution of federal electoral districts and the allotment provided to cities.[137]

The courts have been clear that section 3 does not protect the right to vote or to stand as a candidate in municipal elections,[138] or indeed in elections other than federal or provincial ones.[139] Those doctrines are the result of interpretive decisions by courts that are certainly plausible interpretations of the text of the Charter, but they were choices made by judges among many alternatives. Municipal elections must still comply with freedom of political expression in section 2(b) as well as equality rights in section 15.

While I do not challenge here the overall constitutional status of municipalities, in my view their existence as creatures of the provinces does not preclude the Constitution from ensuring fair procedures and meaningful participation in municipal elections, given their important role as democratic orders of government. It is clear that municipal elections matter in a way that they did not in 1867. In the absence of a role for section 3, section 15 provides some protection, as it would bar discrimination against specific voters in election law. Political expression could fill in gaps as well, if section 2(b) treated voting as expressive activity and operated as a functional substitute for section 3 when applied to municipal elections. There is no controversy that section 2(b) applies to municipal elections. The issue is simply one of interpretation of the scope of the section 2(b) doctrine. If one accepts that the unwritten principle informs the interpretation of section 2(b) of the Charter, it follows that the provision guarantees fair procedures in municipal elections.

The Court of Appeal for Ontario, however, threw cold water on the prospect of constitutional protections for municipal elections by adopting an unnecessarily narrow interpretation of section 2(b). The issue came to a head in relation to the Ontario Legislative Assembly’s Bill 5, which proposed to reduce the number of municipal wards for the City of Toronto.[140] The principled justification for the legislation was to ensure the number of wards and their boundaries were the same as for provincial and federal electoral districts.[141] There is a legitimate argument to be made that having consistent electoral boundaries across different orders of government has benefits for representation.[142] Undermining this claim of enhanced representation was the reality that the federal map copied for provincial districts and then Toronto’s wards departed dramatically from the notion of “one person, one vote” or representation by population.[143] The implication is that a federal/provincial map that was potentially unconstitutional itself was being imported into Toronto.

The complicating factor at the heart of the litigation is that the province imposed the changes in the middle of the city’s election campaign.[144] This change in the boundaries and reduction in the number of municipal wards mid-campaign was unprecedented. The boundaries of wards or districts should be set well in advance of elections—as they usually are in Canada—whether for federal, provincial, or municipal votes.

The application judge, Justice Edward Belobaba, garnered headlines for striking down the provisions in the law that altered the number of wards. He did so on the basis of section 2(b). He controversially imported the guarantee of “effective representation”[145] from section 3 of the Charter into section 2(b).[146] He found two breaches of section 2(b) where Bill 5 unduly diminished effective representation. First, he concluded that interference with the basic structure of the election mid-campaign harmed voters and candidates.[147] Second, he found that it was also a breach to increase the population of each ward, which was the direct result of decreasing the number of wards in the city.[148] In his holding, the population increase harmed voters because it hindered their effective representation, given the larger demands on their representatives in a more populous ward.

The Court of Appeal on a stay motion pending the appeal of Justice Belobaba’s ruling expressed clear disapproval of his reasoning. A majority of the Court ultimately overruled him on the merits,[149] though with a dissent by Justice MacPherson based on freedom of expression. The majority of the Court of Appeal concluded that Justice Belobaba “wrongly imports the content of s. 3 into s. 2(b) in order to circumvent the decision of the constitutional framers not to extend the protection of s. 3 to municipal elections.”[150]

Justice Belobaba was incorrect in my view in finding a second breach related to increases in the population of each ward, though not because it imported the section 3 doctrine into section 2(b). His conclusion that there was a second breach recognized constitutional damage caused by the presence of populous districts. The notion that populous districts harm their constituents is unknown in Canadian constitutional law, including in section 3 or elsewhere.[151] To the best of my knowledge, it is not a relevant concern in other comparable jurisdictions either.[152] The relative population of districts matters, not their absolute size.[153] A claim that more populous districts harm representation necessarily implies keeping the population of wards relatively constant. Such an outcome can only be achieved by endlessly increasing the number of members in the legislative body to keep pace with population growth. A larger legislative body has potentially negative consequences for representation.[154] The majority of the Court of Appeal was right to overrule Justice Belobaba on this particular point, though it did so for the wrong reasons.[155]

The first breach found by Justice Belobaba in relation to the harm caused by mid-election changes has more merit. Interpretation of the relevant Charter provisions directly raises the role of the unwritten principle of democracy in constitutional interpretation. Electoral districts are fundamental aspects of conducting elections. They shape representation, candidate strategy and communications, informal coalitions among representatives in a jurisdiction without political parties, and the diversity of representatives, among other matters.

The reasoning by the majority of the Court of Appeal on the impact of the mid-campaign interference leaves much to be desired. Municipalities are creatures of the provinces, they held, and if a province can create or eliminate a municipality, then it can legislate as to how municipal elections are to proceed. The province can change the rules mid-campaign, just as it could simply stop the practice of having municipal elections. In other words, there is no requirement of fair procedures in the election.[156] The consequences of this approach are untenable. The implication of the Court of Appeal’s conclusion is that voters in Toronto could select a mayor, but then the premier could appoint the second-place finisher as the winner or simply declare that he would not respect the outcome. Even on the facts of the case, mid-election interference was a serious and unprecedented action. Imposing mid-election changes in the basic rules of the game is qualitatively different from amending legislation about municipal elections in the normal course of things outside of a campaign under the jurisdiction provided by section 92(8) of the Constitution Act, 1867.

The core of the majority’s analysis focused on its disapproval of Justice Belobaba’s approach to the relationship between sections 2(b) and 3 of the Charter. The alternative interpretive approach taken by the majority of the Court of Appeal on the relationship between different rights and freedoms guaranteed by the Charter is a novel one unsupported by the case law. The majority labels the application of the effective representation standard developed initially under section 3 to municipal elections through section 2(b) as a “workaround.”[157] The assertion is that it runs counter to the text of the Charter to protect municipal elections in the same way as federal and provincial elections. There is a superficial logic to this point. Section 3 refers to “elections to the House of Commons or of a legislative assembly.” It does not specifically refer to municipal elections and, therefore, the majority concludes that doctrines developed under section 3 cannot legitimately be used in interpretation of section 2(b) of the Charter.[158]

It is not “circumvention” or a “workaround” for judges to amend doctrines in light of unexpected new scenarios, however, including the unprecedented scenario of mid-election changes in electoral districts. There is no principled reason why Charter rights and content must be interpreted to have unique rather than overlapping content. Justice Miller for the majority acknowledged as much in writing that “[r]ights protections often overlap”[159] and that “the coverage of particular rights can overlap.”[160] Using “effective representation” as a standard initially developed pursuant to section 3 under the auspices of section 2(b) was, in his view, impermissible because it violated the “basic structure” of the Charter, “subsumed” the right to vote within freedom of political expression, and “inflated [the] content” of section 2(b).[161] The only support for this conclusion in the case law that is provided is a reference to paragraphs 79–80 of Thomson Newspapers Co. v. Canada (AG)[162] and paragraph 67 of Harper.[163] These paragraphs require careful scrutiny.

Thomson Newspapers does indeed state that sections 3 and 2(b) are “distinct,” but it did so in the scenario where both might conceivably apply. The paragraphs cited by Justice Miller simply set out that where the two provisions “overlap or come into conflict,” they are to be balanced rather than placed in any hierarchy.[164] It is an assertion of the now familiar approach by the Court that balancing is preferred to conflict in interpreting related rights.[165] They “must [each] be given effect”[166] when both are operative and need to be balanced. The case does not stand for the proposition that the right to vote and freedom of political expression are barred from having similar content where only one is operative. The passage cited from Harper relies on Thomson Newspapers for the conclusion that the provisions of the Charter are “distinct” and then holds that it is wrong to equate sections 3 and 2(b).[167] These statements in Harper, however, are again in the service of establishing the need to engage in balancing between rights and freedoms.[168] They do not establish a rule against overlapping content where only one right or freedom is in play.

A cleaner solution in my view would have been to rely on the unwritten principle of democracy as the interpretive guide in assessing whether section 2(b) prevents changing the rules in the midst of a municipal election in a manner that affects expressive freedom. This assessment is made independently of what analysis might proceed under section 3. If the “basic structure” of the Constitution is at stake, surely the principle of democracy is relevant. The Court of Appeal majority[169] and dissent[170] rejected the argument that the unwritten principle of democracy barred the mid-election dissolution of the electoral map. The reasoning of the majority is that unwritten principles should not on their own be the basis for invalidating legislation.[171] The notion that the democracy principle could inform the evolution of section 2(b) doctrine is not considered.[172] There is no analysis of the content of the principle as set out in the Secession Reference, as thick or thin, or otherwise.

Justice Miller justifies this approach by rejecting the view that the text has any “gaps” in relation to municipalities or municipal elections. He concludes that “[s]uch gaps are rare. No such gaps exist here.”[173] He reaches this conclusion because section 92(8) of the Constitution Act, 1867 assigns municipalities as creatures of the provinces in the division of powers. The framers turned their minds to municipalities, so it would be wrong for judges to alter the division of powers absent an amendment, in his view. This conclusion may be justified in relation to the division of powers, but it is not persuasive with regard to the Charter. It is true that municipalities are not a “new subject matter.”[174] Yet their importance and role in Canadian democracy have shifted dramatically since 1867 as have legitimate public expectations of standards for the conduct of free and fair elections. The Supreme Court has grappled with the implications of the gaps in the Constitution in relation to municipalities in other contexts precisely because of the ill fit between the text of section 92(8) and the actual role of cities.[175]

The minimum conditions for democracy found in the unwritten principle include procedural fairness in the conduct of elections. Section 2(b) should be interpreted in that light. Such a role for the unwritten principle does not mean the untrammelled implementation of judicial preferences or arbitrary reliance on an elaborate or idiosyncratic theory of democracy. Courts should not require the creation of municipalities or the existence of municipal elections. As Justice Miller correctly points out, doing so would entail wholesale rewriting of the division of powers in a manner that would only be proper by formal constitutional amendment pursuant to Part V. No great disservice was done to the constitutional order by Justice Belobaba’s holding, however, and, in fact, its consequences were quite narrow. It merely required that the province wait to change the ward boundaries until after the Toronto election was over. The legislative sovereignty and authority of the province under section 92(8) was undisturbed.

Interrupting an election mid-campaign to change the rules of the game, including the electoral districts upon which candidates have crafted their campaigns and voters will have their preferences channelled, should be seen as a clear violation of section 2(b). Changing the number and shape of the wards mid-election had a non-trivial impact on political expression. The events in Toronto were harmful to meaningful participation and accountability. Even on a thin, procedural account of democracy, no election is fair if conducted under such conditions. While the drafters of the Charter may not have anticipated such circumstances as those involving the Toronto municipal election, there is nothing improper about interpreting it in a way that is consistent with a procedural account of the unwritten principle of democracy.


This article has sought to advance content for the unwritten principle of democracy that is narrow enough to ensure legitimacy and does not intrude on the purview of other principles, but which provides real constraints on government action that harms meaningful participation. In my view, the unwritten principles are likely to remain an important feature of the Canadian constitutional landscape. The gaps in the existing text, the need for reasoning to occur behind the text in order to ascertain its content, the Supreme Court’s fondness for structural reasoning, and the reality that the onerous amendment procedures frustrate most attempts at constitutional change all point toward the continuing usefulness of the unwritten principles. If this assumption is correct, then the content of the unwritten principles will continue to be fundamental issues for the Canadian constitutional order.

The objections to the use of unwritten principles should be taken seriously. The legitimacy and interrelatedness critiques should give us pause in assessing how, when, and why unwritten principles are used in constitutional interpretation. Those critiques, however, can be addressed by defining the content of unwritten principles in a manner that is consistent with the separation of powers and the judicial role.

A thin, procedural account centred on meaningful participation is one promising way forward for the principle of democracy. Thicker accounts are incongruent with the role of the courts in the separation of powers. Even on a thin account, however, the democracy principle has an important functional role in ensuring fair elections wherever there is a democratic event.

Referendums and municipal elections are two clear areas where the unwritten principle of democracy can usefully inform the interpretation of the relevant doctrine. There is no controversy that sections 2(b) and 15 of the Charter apply to these democratic events. The issue is simply one of setting the scope of the doctrine under those provisions. The unwritten principle should inform and guide the evolution of judicial doctrine on those provisions of the Charter.