The Senate Reference did not provide an ideal situation for clarifying the nature and limits of constitutional change in Canada. The facts gave the Supreme Court little choice other than to recognize the fundamental role that the Senate plays in the Canadian constitutional order and, therefore, to place some of its main features outside the scope of unilateral action by the federal government under its exclusive authority to “make laws amending the Constitution of Canada in relation to the executive government of Canada or the Senate and House of Commons.” Perhaps more controversially, in order to prevent the implementation of consultative elections that, according to the judges, would fundamentally alter the architecture of the constitution, the Court was forced to construe the prospect of a democratically legitimate Senate in a negative light. This construction seems a somewhat perverse and unappealing way to proceed, especially if the constitution is intended to promote, not stymie, democratic improvement.
Of course, the democratic deficiencies of the Canadian Senate (for example, non-elected chamber, non-proportional representation of the provinces, long appointments, historical property qualifications, et cetera) are not unique; they have been widely shared by upper houses in other jurisdictions. Like the Supreme Court, modern and contemporary defenders of bicameralism have nevertheless attempted (with varying degrees of success) to present those democratic deficiencies as strengths: upper houses slow down the legislative process, avoid sudden legislative changes, protect otherwise potentially underrepresented minorities, force legislators to have second thoughts, and so on. Despite these efforts, many jurisdictions during the twentieth century radically reformed their upper houses or moved toward different forms of unicameralism, and in a number of countries (such as the United Kingdom and Ireland), discussions of this nature continue. Viewed in this light, Canada might be considered to be on a backward slide toward a less, not more, democratic scheme of government.
In this essay, we provide both an explanation for the existing state of affairs and a basis for doing something about it. The main challenge is to think in a less blinkered way about what counts as democratic in regard to the role and composition of the Senate as well as the larger constitutional processes at play. In so doing, it might become possible to diagnose the causes of the present malaise and to offer a more satisfying prognosis for a future and healthier role, if any, for the Senate in the Canadian body politic. A commitment to democracy is professed to be the bedrock of Canada’s constitutional arrangement. Our proposal takes the commitment more seriously than most commentators or scholars do. So reinvigorated, Canadian democracy might be better placed to improve, if not resolve, its long-running drama over the Senate’s continuing existence and performance.
The paper is organized as follows. Part I considers the current state of the debate after the Supreme Court’s decision in the Senate Reference. Part II compares the Canadian approach to constitutional amendments to arrangements and practices present in other jurisdictions. Part III builds on those comparative insights and argues that constitutions with similar amendment rules tend to use those rules to protect and promote basic democratic principles. While Canada’s constitution does not explicitly adopt this approach, the Supreme Court has followed a similar route through the recognition of democracy as one of the underlying principles on which the constitution rests. In that sense, the decision in the Senate Reference may have underplayed the role that democracy plays in the Canadian constitutional order. Part IV identifies what we consider to be the main problematic features in the structure and operation of the current Senate. In Part V, we advance a substantive proposal that would address those problems: reconceiving the Senate as constitutional reviewer.
- The Impasse
The Senate is the upper, but less influential, chamber of Canada’s bicameral Parliament. It has been the subject of countless plans and reform proposals about its membership, its appointment process, its powers, and its democratic legitimacy. While these transformative efforts have been initiated and motivated by a diverse range of vested interests and political ideologies, Canada’s Senate remains largely unchanged in its role or structure. Consequently, while there is an apparent consensus in support of the need for change, there is a complete lack of agreement on the direction that any change should take and the process by which it might be achieved. All in all, Canada’s Senate remains unreformed because it is neither sufficiently loved nor sufficiently unloved; it is tolerated, but not embraced.
Efforts to change or abolish the Senate have been a staple feature of Canada’s political and constitutional landscape since even before its creation in 1867. As far back as the 1840s, when the Senate was the Legislative Council of the Province of Canada, politicians and pundits have locked horns over its role and its composition. Indeed, in 1885, the ruling Liberal Party adopted, but failed to implement, a policy demanding that senators be elected. For the next hundred years or so, debate spluttered and surged over the nature of the appointment process and whether the Senate should simply be abolished. However, in the late 1970s and early 1980s, the debate over the possibility of electing senators was rekindled; Trudeau’s efforts to implement a national energy policy alienated the western provinces and galvanized them to push for a so-called Triple-E Senate – elected (by local citizens), equal (between provinces), and effective (having genuine power).
Following the failed Charlottetown Accord in the early 1990s, the Triple-E initiative lost its lustre, but the federal Conservative government persevered with the idea of a reformed Senate, especially the plan to elect senatorial candidates. However, not only did this run into strong opposition from both the New Democratic Party and the Bloc Québécois, but pressing questions were raised generally about the constitutionality of any proposed reforms. Whatever the Harper government’s motives, a reference was presented to the Supreme Court to clarify the constitutional status of different proposals. The case was heard in late 2013 and a judgment was rendered at the end of April 2014. In short, the Supreme Court nixed the realistic prospect of any Senate reform by the federal government without provincial sign-on. Barring an extraordinary degree of federal–provincial consent to push through institutional reforms or to amend the constitutional provisions for amendment, Senate reform is now considered off the political agenda for the immediate future.
The basic thrust of the Supreme Court’s decision in the Senate Reference is straightforward. In its unanimous decision, the Court decided that, although the Senate is a federal institution, any significant changes to it—tenure, consultative elections, abolition, et cetera—cannot be achieved by the federal government through section 44 of the Constitution Act, 1982, which allows the Canadian Parliament to unilaterally amend the Constitution of Canada. “The Senate,” stated the Court, “is one of Canada’s foundational political institutions. It lies at the heart of the agreements that gave birth to the Canadian federation.” Consequently, as changes to the Senate “engage the interests of the provinces,” such changes must be approved in line with the constitutional requirements for change that demand, to varying degrees, provincial participation and approval. Outright abolition, the most radical change, requires unanimous federal–provincial consent, because to allow otherwise would be to flaunt and “alter our constitutional architecture … by removing the bicameral form of government that gives shape to the Constitution Act, 1867.”
While this outcome has significant constitutional logic and authority, its democratic justification is uncertain and mixed. On the one hand, the decision can be seen as advancing democracy by ensuring that important and democratically represented interests (i.e., those of the provinces) cannot be affected without their political involvement and formal consent. On the other hand, it can be seen as thwarting the democratic process by preventing a duly elected federal government from changing its own institutions in order to make them more democratic and representative. Of course, the role and position of the Supreme Court itself can be viewed in similar mixed terms, both as an integral feature of Canada’s mode of democratic governance, and as an unrepresentative and unelected institution within that governmental structure.
Against this backdrop, it becomes difficult to make any hard and fast criticisms of the Supreme Court’s decision or to defend viable recommendations for major Senate reform in the name of democratic advancement in the Canadian polity. One route might be to take seriously the Supreme Court’s observations about the Senate’s fundamental nature and “role as a complementary legislative chamber of sober second thought.” However, instead of attributing to the Senate the task of evaluating the desirability of the policies advanced by particular bills, we propose to reconceive it as an entity called to reflect (in an attitude of sober second thought) on the constitutionality of those bills. This proposal might be thought of as both a too limited (for supporters) and too expansive (for abolitionists) understanding of the Senate’s possibilities and powers. Nevertheless, when advocated in the name of enhanced democratic governance, this understanding can serve as the basis for the Senate becoming a fuller and more empowered partner in federal government.
It must be emphasized to those democratic purists, be they populist reformers or abolitionists—who maintain that any change must satisfy the most demanding tenets of the democratic ideal—that the best should not be the enemy of the good. While an elected Senate might well be the best democratic circumstance or outcome (if there is to be any Senate at all), there is no reason why important recalibrations cannot be made within the existing constitutional arrangements and parameters. This is a decidedly second-best solution. But there is no all-or-nothing dilemma despite the heated protestations of critics of the Supreme Court’s Senate Reference. Smaller and less ideal changes can be effected that might advance the democratic agenda more than resigned or bitter acquiescence to the status quo. As with any effort at political change, it is important to start where things presently stand and seek change from there: An exasperated resignation to the status quo or a high-minded but naïve idealism is no substitute for a principled pragmatism.
As noted earlier, rather than attack or defend bicameralism, we will argue in favour of attributing to a democratically reconstituted Senate the primary responsibility of reviewing the constitutionality of legislation. Such an approach, we suggest, would serve at least two main functions that would augment the overall democratic legitimacy of the constitutional order. First, it would increase deliberation by elected officials on constitutional issues, potentially decreasing the instances of judges striking down legislation. Second, it would reduce the political costs of using the override provisions of the Charter by making clear that there is a disagreement on the meaning of rights, rather than a rights-violating legislature overriding the decisions of a rights-protecting Court. Before exploring that approach in more detail, we will first look at the constitutional processes at play in the Senate reform debate from a comparative perspective.
II. Of Tiered Amendment Rules
The Constitution Act, 1982 does not have a single rule for constitutional amendments. This is not unique to Canada, but is a feature common to many constitutions. For example, Article V of the Constitution of the United States provides at least four methods to produce a valid constitutional amendment. According to the “least demanding” one, two-thirds of the members of each house of Congress may propose an amendment, which would become valid upon ratification by the legislatures of three-fourths of the states. The most demanding method requires amendments to be proposed by a special national convention called by Congress on application of two-thirds of the states and ratified by special conventions on three-fourths of the states. Not surprisingly, this method has never been used.
With the only exception being the rule that depriving a state from equal representation on the Senate requires the consent of that state (which could be seen as involving a fifth method for amending the constitution), those multiple processes are content-independent: each of them can be used to make any textual change in the 1787 Constitution regardless of how fundamental that change is. Some United States constitutional lawyers, however, have argued that the substance of an amendment has Article V implications. For example, William Harris maintains that in order to correctly amend the US Constitution, one must first examine the substance of the amendment to be adopted, and then determine which of the methods provided by Article V is to be used. In Harris’s view, if an amendment fundamentally alters the way in which political power is exercised, it must be adopted by one of the most demanding methods of change (which, in the US constitutional imagery, are seen as having a closer connection to the people).
While one cannot find textual support for Harris’s position in the US Constitution, the amendment rule contained in the Canadian Constitution Act, 1982 is in a certain way consistent with his view. The Constitution Act, 1982 establishes five different amendment methods. It is premised on the idea that the more fundamental a change is, the more stringent the requirements of the method used to produce it should be. For example, the Constitution Act, 1982 identifies a number of changes (like those pertaining to the office of the Queen, to the use of the English or French languages, and to the composition of the Supreme Court) that must be adopted by the strictest procedure possible (i.e., the unanimous consent of federal and provincial legislatures). Such a level of agreement would arguably indicate a higher degree of popular support. Other changes of a less fundamental nature may be adopted by less demanding mechanisms (i.e., mechanisms requiring a lesser degree of federal–provincial consent).
This approach is not unique either. The Constitution of Spain, adopted in 1978, also exhibits a tiered amendment rule. Most amendments to the constitutional text can be achieved through the general procedure established in Article 167, which places the amending power in a supermajority of both legislative houses. If requested by a tenth of the members of one of the chambers, ratification by referendum is necessary. Nevertheless, changes that involve the “total revision” of the constitution, or that affect, for example, the form of government (i.e., the monarchy), the official language, the indivisibility of the Spanish nation, or recognized rights and liberties, must be first proposed by two-thirds of both chambers. The chambers are then dissolved, a new election called, and the newly elected legislators must approve the change by a two-thirds majority. Finally, the change in question comes into effect only if ratified by the people in a referendum.
The Canadian and Spanish amendment rules protect what has frequently been termed the constitutions’ basic structure by making the adoption of certain amendments particularly difficult. Some courts in other jurisdictions have attempted to create an even stronger protection for those basic constitutional elements by making them legally unamendable. For example, in Kesavananda Bharati v. Kerala, a case that dealt with a series of land reforms that affected property rights, the Supreme Court of India famously determined that, while Parliament had the power to amend any constitutional provision, it could not alter the basic structure of the constitution. According to the different concurring opinions, the basic structure included the principle of constitutional supremacy, the republican form of government, federalism, the welfare state, individual liberty, and secularism. Different versions of this doctrine have been adopted by courts in countries as dissimilar as Colombia and Belize.
III. Of Basic Structures and Democracy
A constitution’s basic structure is the legal manifestation of the fundamental commitments on which the constitutional regime rests. It is thus not surprising that Canada, Spain, and India, as constitutional democracies, have similar ideas about what constitutes their constitutions’ basic structure. However, Spain and India, as well as other jurisdictions that have more recently adopted constitutions with tiered amendment formulas, such as Bolivia and Ecuador, have included as part of their constitutions’ basic structure—either through judicial decisions or through explicit constitutional provisions—the protection of fundamental rights. In those countries, amendments that tend to abolish fundamental rights are extremely difficult to bring into existence, or simply inadmissible. Since respect for many of those rights is generally understood as a precondition for democracy, the purpose of that approach is best understood as an attempt to prevent democracy from being weakened through formal constitutional change.
Put differently, these legal orders treat democracy as an integral part of the constitution’s basic structure. Any attempt to debilitate this arrangement through a change in the constitutional text would be as serious (from a constitutional perspective) as an alteration of a federal system or of a monarchical or republican form of government. The amendment rules contained in the Constitution Act, 1982 do not formally protect democracy in this way. However, there are some textual indications of the fundamental role that democracy plays in the Canadian constitutional order, such as the fact that the rights listed in sections 3 to 5 of the Charter (under the heading “Democratic Rights”) are placed out of the scope of the legislative override provision. Perhaps more importantly, the Supreme Court of Canada has recognized democracy (together with federalism, respect for minorities, and constitutionalism and the rule of law) as one of the underlying principles that form part of the polity’s basic constitutional structure.
In the Secession Reference, the Court stated that “it would be impossible to conceive of our constitutional structure without [these principles]. [They] dictate major elements of the architecture of the constitution itself and are as such its lifeblood.” Taking into account those principles, the Court must interpret the constitution “with a view to discerning the structure of government that it seeks to implement.” On the principle of democracy, the Court expressed that it “has always informed the design of our constitutional structure, and continues to act as an essential interpretive consideration to this day” and “[i]n institutional terms, democracy means that each of the provincial legislatures and the federal Parliament is elected by popular franchise.” Of course, democracy will sometimes need to be understood in light of the other underlying principles, as in the Secession Reference with respect to the principle of federalism.
In light of this approach, it would be expected that the extent to which a proposed amendment runs contrary to one or more of those underlying principles (without at the same time promoting another) would have an effect on the way the Court interprets the amendment rule. For example, in the event of doubt as to the applicable amendment process, the adoption of an amendment that alters Canada’s federal structure would likely have to take place through one of the most stringent procedures. On the contrary, if a proposed amendment would promote one of those principles (again, without negatively affecting another), one would expect the court to err on the side of making the change more easily achievable. This would not seem to be a controversial position to take, since a constitutional order that rests on a number of underlying principles would always be presumed to be striving toward them. In terms of Senate reform, this is particularly relevant for the proposal about consultative elections; the word “consultative” should be stressed, as it must be remembered that what was at issue was not the prospect of real democratic elections, but a system in which senatorial nominees would be endorsed by the populations of the provinces that decided to hold elections and in which the Prime Minister would remain able to ignore those nominees when making a recommendation for appointment.
As the Supreme Court recognized, such an arrangement would not require a formal change in the constitutional text.  The Governor General would still appoint senators according to section 24 of the Constitution Act, 1867. It would nevertheless indirectly affect the convention that the Prime Minister recommends Senate nominees to the Governor General because the former would be required to consider—although not being bound by—the result of provincial elections. However, such a constitutional reality (i.e., non-binding provincial elections for senatorial candidates along with a situation in which the Prime Minister takes into account, without necessarily following, the result of those elections) could in any case operate as an informal practice or understanding as to the ways in which Senators should be appointed; this could happen even in the absence of federal legislation. If such a practice (which may evolve into a convention) existed, it would have to be tolerated by the courts because it would not involve any taint of unconstitutionality.
For example, it would be constitutionally valid for one province (or for a number of provinces) to legislate so as to create a system of non-binding elections for senatorial candidates and communicate to the federal government the election results (as has been the law in Alberta since 1989). In light of this, it is difficult to understand why a similar state of affairs cannot be achieved through federal legislation. And it is not that informal practices that may evolve into conventions cannot be subject to judicial invalidation. For example, there could never be a convention that the federal government will not allow members of a particular minority to exercise the right to vote. Any attempt by the government to engage in such a practice in the first place would be invalidated by the courts, because the practice itself (however informal) would be unconstitutional. Naturally, a piece of legislation that attempts to achieve the same result would be invalid as well. But if a particular governmental practice is constitutional, then how could it be unconstitutional to transform it into law through section 44 of the Constitution Act, 1982?
Moreover, consultative elections would arguably promote the principle of democracy without significantly affecting any of the other underlying principles of the Canadian constitutional order. The Prime Minister, for example, would be free to ignore the provincial election of senatorial candidates that have shown disrespect for minority rights, and the role that the Senate plays within the federal governmental structure would be maintained (for example, provincial interests would still be represented in the same proportion). In this respect, it is constitutionally odd that the Court rendered the establishment of a system of consultative elections that would increase the democratic legitimacy of a government institution so difficult to achieve. From a comparative constitutional law perspective, it may even be seen as a constitutional aberration. In fact, if it had been the other way around (i.e., if the Canadian Senate was already an elected body and Parliament had attempted to transform it into a non-elected entity), the Supreme Court not only would have been justified in reading the amendment provisions of the Constitution Act, 1982 as requiring the most demanding process (i.e., federal-provincial unanimity), but perhaps to disallow the amendment altogether (deciding, as other jurisdictions have done, that the constitution has a core that can only be changed through an exercise of the original constituent power).
IV. Upper Houses and Democratic Legitimacy
Upper houses have a bad reputation among democrats. Historically, they have been assigned the role of preventing government from degenerating into a system that protects the interests of the many over those of the few. Upper houses were usually created to play a balancing role; they were conceived as an aristocratic element that would ensure that the system did not turn into an absolute democracy. Accordingly, their members invariably belonged to the upper classes of society; this explains the names “upper house” and “lower house.” They were seen as providing members of the propertied elites with the opportunity to protect their own interests, which frequently ran contrary to those of the general population. In fourteenth-century England, for example, the division between the House of Lords and the House of Commons was based on a distinction “between those who represented themselves and those who represented their communities.”
Nowadays that approach has been largely abandoned. Instead, defenders of bicameralism usually argue that the advantage of an upper house is that it “slows down the legislative process, renders abrupt change difficult, forces myopic legislators to have second thoughts, and thereby minimizes arbitrariness and injustice in governmental action.” But the old, elite-based understanding is still reflected in the form that bicameralism assumes in jurisdictions that have retained largely unreformed upper houses. For example, it is still common to find unelected, or partially unelected, upper houses, such as the Canadian Senate, the Irish Seanad, and the British House of Lords. Nevertheless, most upper houses are now elected, even if they were originally composed of appointed members. This process toward the democratization of upper houses is exemplified by the reform of Legislative Councils in Australian states during the twentieth century.
Partly because of their dubious democratic legitimacy, upper houses are generally less prestigious and powerful than their elected counterparts. However, the fact that they lack democratic legitimacy is sometimes seen as a virtue rather than a vice. For example, Mughan and Patterson maintain that “for parliamentary governments to create directly elected upper houses would be potentially to create a whip for their own backs.” “Direct election,” they argue, “would endow senates with the same democratic legitimacy as lower houses and, especially when controlled by a different partisan majority, would turn them into a competitor” for the political authority of lower chambers. Or, in the words of the Supreme Court of Canada: “[t]he proposed consultative elections … would weaken the Senate’s role of sober second thought and would give it the democratic legitimacy to systematically block the House of Commons, contrary to its constitutional design.”
While the upshot of this view is that, given their lack of democratic legitimacy, upper houses would normally be very wary of interfering with the policy decisions of the elected house, the fact is that they sometimes work as a means for “slowing politics down”; they can delay, and sometimes block, the passage of legislation they consider undesirable. There are many historical examples of Conservative upper houses preventing the government of the day from adopting progressive measures, including in the United Kingdom several times during the twentieth century (situations that eventually led to major reform of the powers of the House of Lords). For these and other reasons (including their failure to play any meaningful role in the law-making process), the existence and nature of upper houses has usually been the subject of intense discussions. In some places, including New Zealand and Denmark, those discussions have led to their abolition; in others, such as Canada, France, Ireland, and Spain, they have resulted in different attempts at reform.
It has thus been suggested that upper houses are “essentially contested” institutions, always the subject of dispute: “Many countries choose not to have one, others have them but then do away with them, and still others keep them but are engaged in an apparently incessant dialogue about how they should be reformed.” Canada falls, of course, in the third camp. Under the formal constitutional arrangements in place, the Senate is a full partner in Canada’s bicameral system of federal governance. The only restrictions on the powers of the Senate are twofold. First, under the Constitution Act, 1867, bills that impose taxes or appropriate public funds cannot be introduced in the Senate. Secondly, while the Senate can delay the approval of a constitutional amendment that has been passed by the House of Commons, it cannot veto it indefinitely; the House of Commons can override the Senate’s rejection of a constitutional amendment after a 180-day wait.
However, as a less politically partisan institution, the Senate follows a convention whereby it plays a decidedly secondary role in the legislative process when compared to the more democratically authorized House of Commons. Accordingly, most bills arise in the House of Commons and will receive the approval of the Senate, with only minor amendments or friendly changes. Nevertheless, this state of affairs has not prevented the Senate from taking an active and disapproving role when it thought the situation demanded it. As revealingly described by Sir John A. Macdonald (and reiterated very recently by the Supreme Court in the Senate Reference), the Senate functions as a body of “sober second-thought” that is intended to curb the democratic excesses of the House of Commons. Even though Macdonald also maintained that the Senate “will never set itself in opposition against the deliberate and understood wishes of the people,” the fact is that every now and then, the Senate will challenge the decisions of the democratically elected House of Commons.
In its early years up to the mid-1920s, the Senate effectively scuppered government legislation on around 180 occasions. In the next sixty years or so up to 1987, the rate of rejection decreased and only about forty-five or so of government bills in total were turned back. Since then, the Senate has occasionally inserted itself into high-profile and controversial government initiatives. For instance, in the late 1980s, it refused to pass the Canada–United States free trade bill and later the goods and services tax legislation. On both occasions, the government managed to force through the legislation, but only after a federal election in one instance and the appointment of extra senators in another. Again, in the early 1990s, the Senate narrowly turned back government efforts to enact new abortion legislation following the constitutional invalidation of earlier legislation by the Supreme Court. More recently, in 2010, the Senate refused to pass a government bill intended to regulate greenhouse gases.
Of course, these more vigorous and perhaps not “sober” interventions by the Senate have aroused controversy. They have added convenient fuel to the abolitionist fire. Nevertheless, the problem is not only the current Senate’s undemocratic nature (after all, an elected upper house with a different ideological composition to the government of the day that insists on blocking its policy decisions would probably raise a good amount of discomfort, too), but the fact that it is conceived as a “legislative body of sober second thought”; that is to say, as an entity whose main function is to engage in what the Supreme Court recently identified as “legislative review.” Such an entity would frequently find itself in the position of disagreeing with the policy decisions of the other legislative chamber, and in some cases, be tempted to prevent those decisions from becoming law. In this sense, and in light of the perceived constitutional impasse created by the Supreme Court’s Senate Reference decision, abandoning the view of the Senate as a legislative chamber seems like a promising alternative.
V. The Senate as a Constitutional Reviewer
The approach that we take to Senate reform (and to any other kind of legal or political reform) is what might be called “democratic pragmatism.” This approach is built on the premise that democracy is better understood as being not a ground or foundation for politics but a dynamic process through which politics is experienced and evaluated; the institutional forms and social practices of democracy must themselves be open to debate and revision in line with the experience of society’s participating members. Such a strong approach to democracy is very strategic and political. Rather than talk in absolute terms about the overall democratic quality of society, it is preferable to concentrate on those measures that might make a society more democratic. Viewed in this way, democracy is not a black-and-white idea or practice; a society is not either democratic or undemocratic. It is a matter of shading and degree. While it will be necessary to make trade-offs within a democratic society, it is unnecessary and ill-advised to make trade-offs with democracy.
Consequently, rather than think about democracy as one particular thing or to evaluate a state of affairs as being democratic or not, it is better to ask whether taking a particular step is likely to advance democracy; this will be a contextual and contingent assessment. As such, the type of democratic commitment that we defend is pragmatic and contingent rather than idealistic or absolutist. The primary considerations in any effort to effect change will always be: which present and realizable measures will best increase people’s participation and control over those institutions and practices that most affect their lives today? Or, which present and realizable measures will best decrease the hold that less democratic institutions have over more democratic institutions? It is always a contextual balancing of better or worse, not a final assessment of absolutes and ideals.
While it is better that more citizens be involved more of the time than fewer citizens be involved less of the time, it seems counterproductive to deny the adjective “democratic” to a reform proposal or strategy because it does not involve everyone all the time. If incremental change is all that is viable, then doing something is better than doing nothing and capitulating to the status quo. Such an approach strongly recommends greater efforts to pluralize and circulate the sites and situations in which more democratic—even if imperfect and flawed—interventions are encouraged. By being flexible and pragmatic, democrats can accept that compromise as well as conviction is an important quality of democratic advocacy. For instance, in terms of the constitutional role of the courts, no amount of ingenious finagling can manage to square two incompatible institutional practices—judicial review and popular sovereignty. Indeed, judicial review, as presently conceived, is anathema to the strong democrat; it is an elite practice that might, at best, occasionally be democratic in being for the people, but it can never be by the people (or else it would cease to be judicial review).
That said, any reform that moves some power away from the courts and toward other organs of government is to be encouraged. Our proposal should be understood in that vein: assigning the Senate the task of reviewing the constitutionality of bills proposed by the House of Commons with the purpose of ameliorating the democratic deficit created by the institution of judicial review of legislation, and therefore contributing to the overall democratic legitimacy of the constitutional order. In so doing, the Senate would be acting as a chamber of “sober second thought,” not as to the desirability of the policies advanced in the relevant bills, but as to their consistency with the constitution. Of course, the line between law and policy is blurred. However, the important point is that the Senate should view itself as engaged in a traditional judicial task.
Such a proposal is not novel. In fact, upper houses in other jurisdictions have also been assigned the task of preventing the adoption of unconstitutional bills. For example, in France and Poland, when members of the upper houses believe that a bill is unconstitutional, they can seek a ruling from the relevant court. In Canada, the Senate has already engaged in some forms of constitutional review, as occurred with the “Clarity Bill” in the year 2000.
We envision such a system as having the following features, which in time might evolve into constitutional conventions. First, bills should be introduced in the House of Commons, not in the Senate. This is generally the case today, but in a system in which the Senate’s role would be limited to that of reviewing the constitutionality of legislation, it is important that senators not see themselves as lawmakers. Second, the federal government would make clear to potential Senate appointees that they should understand their function to be that of constitutional, not legislative, reviewers. That is to say, they are not to second-guess the policy decisions of the lower house, but to deliberate on the constitutionality of legislation and, if necessary, to formally communicate to the House their conviction that a particular bill appears to be contrary to a constitutional provision. Third, attempts should be made to compel potential senators to voluntarily step down from office after a relatively short term (for example, ten years).
Fourth, even in cases in which there is a disagreement between the Senate and the House as to the constitutionality of a bill, the former body should not exercise a veto power, except perhaps in extraordinary circumstances (for example, a House of Commons that approves a bill intended to result in gross human rights violations). In light of a disagreement between an elected and a formally non-elected chamber, the view of the elected chamber should prevail. Fifth, if senators are to continue to be appointed by the Governor General on the recommendation of the Prime Minister, provinces that are committed to an elected Senate should put in place a scheme of non-binding elections of senate nominees; the results of these elections would be formally communicated to the federal government (which would then be expected to consider those results when making appointments). A fully appointed Senate with constitutional review functions, after all, would be akin to an enlarged Supreme Court, which, in the words of Jeremy Waldron, “would be the worst possible outcome.” Finally, the Supreme Court would maintain its power to strike down legislation.
In line with the democratic ambition of this proposal, it is important to ensure that the Senate’s deliberations about the constitutionality of federal initiatives are themselves carried out in as thoroughly a democratic way as practicable. To do otherwise would not only undercut the legitimacy of the new role for the Senate, but also confine the structure and substance of such deliberation. The Senate must set its own agenda and develop its own informing theoretical framework in determining what constitutionality means and what it does and does not demand. In particular, it will be essential that Senate debates are not oriented around the Supreme Court’s constitutional jurisprudence or even performed within its capacious argumentative shadow. This would compromise the whole democratic impetus and raison d’être of why the Senate’s institutional responsibilities should be constitutionalized. While Supreme Court jurisprudence might be one resource to be consulted and weighed, it cannot be utilized as the scales of judgment: it must not dominate the discursive orientation and analytical approach of the Senate. This would be to throw good democratic money after bad constitutional money.
For all the democratic flaws of the Senate, its greater involvement in debate and decisions about the constitutionality of federal legislation would be preferable to present arrangements: a slightly democratic authority beats a non-democratic authority. Accordingly, an expanded role for the Senate in the constitutional review of legislation would advance, no matter how incrementally and imperfectly, the commitment to democratic governance. By shifting the institutional fulcrum of constitutional review from the courts to a more democratically reconstituted Senate, it might be possible to augment the overall democratic legitimacy of the constitutional order. Such a reform would address, although by no means eradicate, the democratic failings of a society that persists in placing enormous power in the hands of an unelected and unrepresentative judiciary. To borrow from John Hart Ely, “we may [rant] until we’re blue in the face” that the Senate isn’t “wholly democratic, but that isn’t going to make courts more democratic” than the Senate.
One of the likely outcomes of this arrangement is that if the Senate engages in a robust and good faith deliberation about the constitutionality of a controversial bill, and concludes that it is consistent with the constitution, it is likely that the Supreme Court will be more restrained in using its power to strike down federal legislation (Supreme Court review of provincial legislation would not be directly affected by this proposal). Moreover, this approach might invigorate and render more palatable the resort to the override provisions of the constitution. Although the Constitution Act, 1982 attempted to create a system of constitutional review that would allow Parliament to have the ultimate authority (in most cases) as to the validity of legislation, it is the Supreme Court, another unelected institution, which enjoys the de facto final word.
The fact is that since its introduction in 1982, section 33’s notwithstanding clause has hardly featured in the political operation of the federal or provincial governments. While some provinces have utilized the override provision to insulate legislation from judicial interference, most have not; seven of the ten provinces and two of the three territories have refrained from relying upon it. Most pertinently, the federal government has never exercised its section 33 powers in the more than thirty years of its existence. This state of affairs has become politically accepted on the basis that the use of section 33’s notwithstanding clause is somehow undemocratic and, therefore, illegitimate. From a democratic perspective, this is an entirely troublesome way of looking at matters. Section 33 is as much part of the constitution as any other clause. Indeed, the inclusion of section 33 was essential and instrumental to the adoption of the Charter.
Without section 33, it can be forcefully and validly argued that there might have been no Charter at all and even no patriation of the constitution. Moreover, the upshot of the condemnation of section 33 as somehow undemocratic or even taboo is to entrench further the almost exclusive authority of courts to determine whether the legislative initiatives of an elected Parliament are constitutional. This is tantamount to the elitist tail wagging the democratic dog. A more active constitutional role for the Senate in scrutinizing legislation would have the salutary effect of reducing the political costs of Parliament using the override provision of the Charter. In particular, it would work toward establishing that, when the constitutionality of a piece of legislation is successfully impugned by the courts (after having passed the Senate’s constitutional review), the use of the override would not be a case of a legislature insisting on violating constitutional protections. Instead, it would put the focus on the fact that there was actually a disagreement between two organs of government—legislature and courts—on the meaning and operation of rights.
In light of such explicit disagreement, it would be seen as perfectly reasonable that the organ with the highest degree of democratic legitimacy (the legislature) decide every now and then to exercise its override powers to have the final word. This is a more accurate and compelling understanding of how a democratic mode of constitutional authority should be exercised and implemented. Even among institutions acting in good faith, there will be ample room to agree to disagree over what best promotes and achieves the ambitions of democratic governance in modern society. Giving the Senate a role in scrutinizing legislation for constitutional value would open up that debate and encourage the more enlightened view that what democracy demands is neither fixed nor certain; the vernacular of absolutes is anathema to a democratic polity. As befits its argumentative and participatory aspirations, such a reform of the Senate would enhance the procedural and substantive dimensions of the democratic project.
Importantly, a Senate tasked with constitutional responsibility would not strive to replicate courts and mimic the judicial (and juristic) preoccupation with establishing neutral and non-political bases for defining rights and legitimating their actions. Instead, it might be accepted that the Senate, as with all governing bodies and agencies, will be involved in the same game—delivering substantive answers to concrete problems. In doing so, while no one institution will have a lock on political judgment as to the best thing to do, all institutions’ decisions will be evaluated in terms of the value choices that they make and the contribution that their decisions make to advance substantive democracy in the here and now.
In this essay, we have sought to make the best of a bad job. The Supreme Court’s interpretation of the Canadian constitution has left committed democrats with few genuine options. However, rather than throw up their hands in resignation, it behooves them to take short-term measures that might open up a space for more effective and wide-ranging future interventions. At the heart of our proposal is the basic democratic belief that the appropriate inquiry in a strong democracy is not to ask whether particular policy-making bodies have acted politically and, therefore, improperly, but whether the political choices that they have made serve that society’s democratic agenda. There are no right answers; policies and programs will always be a contested and contestable issue. The most that democrats can do is to ensure that the sites for debate and decision making are as open, multiple, and representative as possible. An effort to constitutionalize the Senate is one place to begin.
 Reference Re Senate Reform, 2014 SCC 32,  1 SCR 704 [Senate Reference].
 Constitution Act, 1982, s 44, being Schedule B to the Canada Act 1982 (UK), 1982, c 11 [Constitution Act, 1982].
 See Senate Reference, supra note 1 at para 54.
 See Bruce Ackerman, “The New Separation of Powers” (2000) 113:3 Harv L Rev 633.
 For a discussion of historical Senate reform proposals, see Robert A MacKay, The Unreformed Senate of Canada, revised ed (Toronto: McClelland and Stewart, 1963).
 Some argue that the objective of an improved Senate may be achieved without constitutional reform. See e.g. Serge Joyal, ed, Protecting Canadian Democracy: The Senate You Never Knew (Montréal: McGill-Queen’s University Press, 2003).
 Senate Reference, supra note 1 at para 1. See also Reference Re Authority of Parliament in Relation to the Upper House (1979),  1 SCR 54, 102 DLR (3d) 1; Reference Re Bill C-7 Concerning the Reform of the Senate, 2013 QCCA 1807, 370 DLR (4th) 711.
 Senate Reference, supra note 1 at para 72.
 See Constitution Act, 1982, supra note 2, s 41(e).
 Senate Reference, supra note 1 at para 97.
 Ibid at para 70.
 Canadian Charter of Rights and Freedoms, s 33, Part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (UK), 1982, c 11 [Charter].
 The other two methods involve two-thirds of the members of each house of Congress proposing an amendment to be ratified by special conventions on three-fourths of the states, and a special national convention called by Congress on application of two-thirds of the states, whose proposals are ratified by the legislatures of three-fourths of the states.
 It has been argued that to the extent that altering one state’s representation in the Senate also deprives all states of “equal Suffrage in the Senate”, the unanimous consent of all states is required for such a change. See Richard Albert, “Constructive Unamendability in Canada and the United States” (2014) 67 SCLR (2d) 181 at 196ff; Sanford Levinson, “The Political Implications of Amending Clauses” (1996) 13:1 Const Commentary 107 at 122.
 Early in the twentieth century, the US Supreme Court rejected the idea that some constitutional changes are so fundamental that they fall outside the scope of Article V. See Leser v Garnett, 258 US 130, 42 S Ct 217 (1922); Coleman v Miller, 307 US 433, 59 S Ct 972 (1939).
 See William F Harris II, The Interpretable Constitution (Baltimore: Johns Hopkins University Press, 1993) at 193–201. See also Elai Katz, “On Amending Constitutions: The Legality and Legitimacy of Constitutional Entrenchment” (1996) 29:2 Colum JL & Soc Probs 251.
 See Constitution Act, 1982, supra note 2, ss 38–49. See also Richard Albert, “The Expressive Function of Constitutional Amendment Rules” (2013) 59:2 McGill LJ 225 at 250.
 For a discussion of the implications of this approach for a democratic conception of constitutionalism, see Joel I Colón-Ríos, Weak Constitutionalism: Democratic Legitimacy and the Question of Constituent Power (Abingdon, UK: Routledge, 2012).
 Kesavananda Bharati v State of Kerala, 4 SCC 225,  60 AIR 1461 (India Sup Ct).
 It is interesting to note that the Colombian Constitutional Court has expressed, in obiter that the abolition of one of the legislative chapters would amount to the creation of a new constitution and therefore is outside the scope of the amending power. See Corte Constitucional [Constitutional Court], Bogatá, 30 July 2008, Sentencia C-757/08 (Colombia).
 See e.g. Carlos Bernal, “Unconstitutional Constitutional Amendments in the Case Study of Colombia: An Analysis of the Justification and Meaning of the Constitutional Replacement Doctrine” (2013) 11:2 Intl J Constitutional L 339; Arif Bulkan, “The Limits of Constitution (Re)-making in the Commonwealth Caribbean: Towards the ‘Perfect Nation’” (2013) 2:1 Can J Human Rights 81. See also Yaniv Roznai & Serkan Yolcu, “An Unconstitutional Constitutional Amendment: The Turkish Perspective: A Comment on the Turkish Constitutional Court’s Headscarf Decision” (2012) 10:1 Intl J Constitutional L 175.
 See Constitución de la República del Ecuador, 20 October 2008 (Ecuador), as amended, art 442; Constitución Política del Estado, 7 February 2009 (Bolivia), art 411.
 The idea that respect for certain rights is necessary for democracy to exist does not necessarily commit one to a particular view of the judicial role. See e.g. Jeremy Waldron, Law and Disagreement (New York: Oxford University Press, 1999).
 See Reference Re Secession of Quebec,  2 SCR 217 at para 49, 161 DLR (4th) 385 [Secession Reference].
 Ibid at para 51.
 Senate Reference, supra note 1 at para 26.
 Secession Reference, supra note 24 at paras 62, 65. See also OPSEU v Ontario (AG),  2 SCR 2, 41 DLR (4th) 1 (in which the court stated that “the basic structure of our Constitution, as established by the Constitution Act, 1867, contemplates the existence of certain political institutions, including freely elected legislative bodies at the federal and provincial levels” at 57).
 See Secession Reference, supra note 24 at paras 66, 88.
 See e.g. the process set out in Bill C-20, An Act to Provide for Consultations with Electors on their Preferences for Appointments to the Senate, 2nd Sess, 39th Parl, 2007 (first reading 13 November 2007) and Bill C-7, An Act Respecting the Selection of Senators and Amending the Constitution Act, 1867 in Respect of Senate Term Limits, 1st Sess, 41st Parl, 2011 (first reading 21 June 2011) [Bill C-7].
 See Senate Reference, supra note 1 at para 52.
 Constitution Act, 1867 (UK), 30 & 31 Vict, c 3, s 24, reprinted in RSC 1985, Appendix II, No 5 [Constitution Act, 1867].
 Contra Mark D Walters, “The Constitutional Form and Reform of the Senate: Thoughts on the Constitutionality of Bill C-7” (2013) 7 JPPL 37.
 For a discussion, see generally Fabien Gélinas & Léonid Sirota, “Constitutional Conventions and Senate Reform” (2013) 5 Revue québécoise de droit constitutionnel 107. See also Patrick J Monahan, Constitutional Law, 2nd ed (Toronto: Irwin Law, 2002) at 488.
 See Senatorial Selection Act, SA 1989, c S-11.5, as re-enacted by RSA 2000, c S-5.
 Or, for that matter, under the general law-making power under section 91 of the Constitution Act, 1867. Perhaps the fact that, at least under Bill C-7, the Prime Minister “must” consider the names of the elected candidates when making a recommendation for appointment (supra note 29, s 31) is what puts consultative senatorial elections outside the scope of section 44. But in the Senate Reference case, the emphasis was not on the language of Bill C-7, but on the fact that after a provincial election takes place, the Prime Minister would be politically obligated to take the results into account (see Senate Reference, supra note 1 at para 62). The same political obligation would arguably exist if provincial elections took place in the absence of federal legislation.
 Bypassing the amendment rule—for example, allowing a constitutional change that clearly requires provincial consent to be adopted unilaterally by the federal government—in the name of democracy would clearly run contrary to the principles of constitutionalism and federalism. But this was not the case in the Senate Reference, as there were reasonable doubts as to the applicable amendment rule—at least with respect to the proposal about consultative elections (see supra note 1 at paras 51–53). It is also possible that some provincial premiers oppose an elected senate because a province-wide elected senator would risk displacing the premier as the province’s voice in Ottawa, but such a political displacement would not directly alter the federal structure of the government.
 See Constitution Act, 1982, supra note 2, s 41. For a discussion, see generally Colón-Ríos, supra note 18.
 See generally George Tsebelis & Jeannette Money, Bicameralism (New York: Cambridge University Press, 1997) at 17–26.
 See Meg Russell, “What are Second Chambers for?” (2001) 54 Parliamentary Affairs 442 at 443.
 It is interesting that in the Senate Reference, the Supreme Court expressed that “the removal of the real property requirement (s 23(3), Constitution Act, 1867) would not alter the fundamental nature and role of the Senate” and therefore could be unilaterally removed by the federal parliament (with a partial exception related to Québec’s electoral law) (Senate Reference, supra note 1 at para 91). The same was stated with respect to the net worth requirement: “There is nothing in the material before us to suggest that removing the net worth requirement would affect the independence of senators or otherwise affect the Senate’s role as a complementary legislative chamber of sober second thought” (ibid at para 88). It is of course true that these qualifications do not play any significant role today, but they certainly did in the nineteenth century and probably for a good part of the twentieth. As in other upper houses around the world, they served as a guarantee that the Senate would be controlled by a “sober” upper propertied class that could serve as a counterpart to the more “passionate” and popular lower house. If the reasoning of the Supreme Court in the Senate Reference had been applied in the late nineteenth century, the removal of the property and net worth requirements would have probably been seen as attracting some of the more stringent amendment formulas. Such a decision would not have been appropriate today because views have since changed. But why can we not also say that views about non-elected senators as a fundamental feature of the upper house have changed too?
 Tsebelis & Money, supra note 38 at 23.
 William H Riker, “The Justification of Bicameralism” (1992) 13:1 Intl Political Science Rev 101 at 101.
 See Tsebelis & Money, supra note 38 at 47.
 See generally Bruce Stone, “Bicameralism and Democracy: The Transformation of Australian State Upper Houses” (2002) 37:2 Australian J Political Science 267.
 Anthony Mughan & Samuel C Patterson, “Senates: A Comparative Perspective” in Samuel C Patterson & Anthony Mughan, eds, Senates: Bicameralism in the Contemporary World (Columbus: Ohio State University Press, 1999) 333 at 339.
 Senate Reference, supra note 1 at para 60. It is worth mentioning that such a view inadvertently provides support not for the democratic decision-making power of the elected commons, but for a political community controlled by the executive. That is to say, an elected upper house may actually serve the purpose of ensuring that other democratic interests are able to act as a check on the decisions of the executive of the day (who will usually have significant control over the majority in the House of Commons). As Waldron has put it in a different context: “I sometimes wonder, when people say how important it is for the House of Commons to retain its ascendancy in our constitution, whether it is really the ascendancy of an elected chamber that they want, or whether they really want to preserve the guaranteed ascendancy of the Crown and its ministers acting through their whip-based domination of the Commons” (“Bicameralism and the Separation of Powers” (2012) 65 Current Leg Probs 31 at 35 [Waldron, “Bicameralism”]).
 Roberto Mangabeira Unger, Democracy Realized: The Progressive Alternative (London: Verso, 1998) at 213.
 See generally Michael Rush, “The House of Lords: The Political Context” in Paul Carmichael & Brice Dickson, eds, The House of Lords: Its Parliamentary and Judicial Roles (Oxford: Hart, 1999) 7. See also Parliament Act, 1911 (UK), 1 & 2 Geo V, c 13; Parliament Act, 1949 (UK) 12, 13 & 14 Geo VI, c 103. For a recent judicial discussion of these Acts, see R (Jackson) v AG (HL(E)),  UKHL 56,  1 AC 262.
 See Russell, supra note 39 at 442.
 Mughan & Patterson, supra note 45 at 338 (for the term “essentially contested,” Mughan and Patterson cite WB Gallie, “Essentially Contested Concepts” (1955–1956) 56 Proceedings of the Aristotelian Society 167).
 See Constitution Act, 1867, supra note 31, s 53.
 See Constitution Act, 1982, supra note 2, s 47.
 For a general discussion, see David C Docherty, “The Canadian Senate: Chamber of Sober Reflection or Loony Cousin Best Not Talked About” (2002) 8:3 J Legislative Studies 27.
 Province of Canada, Legislative Assembly, Parliamentary Debates on the Subject of the Confederation of the British North American Provinces, 3rd Sess, 8th Prov Parl (6 February 1865) at 35–36 (Attorney General John Alexander Macdonald) [Parliamentary Debates]. See also Senate Reference, supra note 1 at para 15 (referring to the upper legislative chamber’s purpose in providing “sober second thought” on legislation adopted by the lower house).
 Parliamentary Debates, supra note 55 at 36.
 Senate Reference, supra note 1 at paras 63, 79 [emphasis added].
 For a fuller defense of this point, see Allan C Hutchinson, The Province of Jurisprudence Democratized (Oxford: Oxford University Press, 2009) at 7–19.
 For a discussion of these mechanisms and of the general role of upper houses as organs of constitutional review, see generally Aisling Reidy & Meg Russell, “Second Chambers as Constitutional Guardians and Protectors of Human Rights” (June 1999), online: UCL School of Public Policy, Constitution Unit <www.ucl.ac.uk/spp/publications/unit-publications/44.pdf >.
 See Docherty, supra note 54 at 36.
 This system may be put into practice without the need for legislation. In fact, if it was to be established through legislation, it is possible that its constitutionality would be put into question by some. Needless to say, one of the odd results of the Senate Reference is that some things that can be informally achieved by public officials (e.g., consultative elections themselves could be established by a set of informal agreements between the federal and provincial governments) cannot be achieved by legislation.
 Waldron, “Bicameralism”, supra note 47 at 50. See also Vernon Bogdanor, “The Problem of the Upper House” in Vernon Bogdanor, Politics and the Constitution: Essays on British Government (Farnham, UK: Ashgate, 1996) 243 at 247: “[i]f one is looking for an upper house with the authority to fulfil the function of constitutional protection … the conclusion seems inescapable that it must be an elected body.”
 This, of course, is the major thrust of the academic literature of “popular constitutionalism.” See Larry Kramer, The People Themselves: Popular Constitutionalism and Judicial Review (Oxford: Oxford University Press, 2004) and Mark Tushnet, “Popular Constitutionalism as Political Law” (2006) 81:3 Chicago-Kent L Rev 991.
 Democracy and Distrust: A Theory of Judicial Review (Cambridge, Mass: Harvard University Press, 1980) at 67.
 Section 33 allows “Parliament or the legislature of a province” to declare “that the Act or a provision thereof shall operate notwithstanding a provision included” in section 2 (Fundamental Rights) or sections 7 to 15 (Legal Rights) of the Charter. The Charter places Democratic Rights (sections 3–5) and Mobility Rights (section 6) outside the scope of section 33.
 For a recent discussion, see Rosalind Dixon, “The Supreme Court of Canada, Charter Dialogue, and Deference” (2009) 47:2 Osgoode Hall LJ 235. The development of a rationale for judicial review that centres on the availability of section 33 seems to be opportunistic at best in light of the fact that it is presently and politically unusable by the federal government. See Peter W Hogg & Alison A Bushell, “The Charter Dialogue Between Courts and Legislatures (Or Perhaps the Charter of Rights Isn’t Such a Bad Thing After All)” (1997) 35:1 Osgoode Hall LJ 75. For a powerful democratic critique, see Andrew Petter, The Politics of the Charter: The Illusive Promise of Constitutional Rights (Toronto: University of Toronto Press, 2010). Also, some commentators have begun to suggest that the failure to resort to section 33 may have rendered its availability constitutionally moot. For a discussion, see Richard Albert, “Constitutional Amendment by Constitutional Desuetude” (2014) 62:3 Am J Comp L 641.
 See Roy Romanow, John Whyte & Howard Leeson, Canada … Notwithstanding: The Making of the Constitution 1976–1982 (Toronto: Carswell/Methuen, 1984).