The theory and doctrine of unconstitutional constitutional amendment in Canada.
Introduction I. Constitutional Amendment in Constitutional Democracies A. The Amendment Power in Constitutional Design (i) Between Flexibility and Permanence (ii) Formal Amendment Rules and Modern Constitutional Democracy B. Formal Prohibitions on Constitutional Amendment (i) Designing Formal Unamendability (ii) Interpreting Formal Unamendability C. Informal Restrictions on Constitutional Amendment (i) Amendment and Revision (ii) The Basic Structure Doctrine II. Constitutional Amendment in Canada A. Formal Amendment in Canada (i) The Escalating Structure of Formal Amendment (ii) The Federalist Structure of Formal Amendment (iii) The Consultative Structure of Formal Amendment B. Judicial Restrictions on Constitutional Amendment (i) Procedural Enforcement (ii) Unwritten Constitutional Principles (iii) Judicial Self-Entrenchment C. Political Restrictions on Constitutional Amendment (i) Parliamentary and Provincial Laws (ii) Convention and Public Expectations (iii) Restrictions by Implication III. A Framework for Judicial Review of Constitutional Amendment A. Procedural Unconstitutionality (i) Subject-Rule Mismatch (ii) Temporal Violations (iii) Processual Irregularity B. Substantive Unconstitutionality (i) Unwritten Fundamental Values (ii) Non-Negotiable Founding Values (iii) Amendment-Revision Unamendability C. Hybrid Forms of Unconstitutionality (i) Statutory Unconstitutionality (ii) The Recognition of Convention (iii) Unconstitutionality by Implication Conclusion
Modern constitutionalism has given rise to a question that has for some time now been the subject of significant scholarly attention: Can a constitutional amendment be unconstitutional? (1) As a normative matter, whether an amendment should ever be declared unconstitutional remains controversial. But in light of contemporary constitutional law and politics around the world, there is no doubt as a descriptive matter that an amendment can indeed be found unconstitutional. The phenomenon of an unconstitutional constitutional amendment traces its political foundations to France and the United States, its doctrinal origins to Germany, and it has migrated in some form to modern constitutional democracies in every corner of the world. (2)
But in Canada the question remains unanswered and indeed largely unexplored. Some of Canada's most formative constitutional controversies have touched the outer edges of the problem, but not squarely enough to generate a theory or doctrine for evaluating the constitutionality of a constitutional amendment. For example, in Reference re Resolution to Amend the Constitution (Patriation Reference), the Supreme Court of Canada concluded that it would have been unconstitutional in the conventional sense for Parliament to pass the patriation resolution without securing substantial provincial consent. (3) Later, in Reference re Secession of Quebec, the Court suggested that negotiations on a formal amendment in connection with provincial secession must respect certain unwritten principles. (4) More recently, the Court advised Parliament that it could not unilaterally make amendments to the method for filling vacancies in the
Senate. (5) These judicial opinions and others intimate that some informal concept of an unconstitutional constitutional amendment has begun to take root in Canada, whether or not it has yet been recognized.
In this article, I suggest a framework for evaluating the constitutional validity of amendments to the Constitution of Canada. (6) It is important to stress that I do not inquire into the legitimacy of invalidating a constitutional amendment. I am concerned instead only with whether and how the Supreme Court of Canada could invalidate a constitutional amendment. (7) I show that although the Court has yet to invalidate an amendment, (8) modern constitutional politics suggest that the Court possesses residual constitutional authority to declare that a future amendment violates either the text or spirit of the Canadian Constitution. This residual authority derives both from the Court's power of judicial review and from contemporary changes to the Constitution "outside" of the Constitution. (9)
Drawing from the judicial review of constitutional amendments around the world, I propose a framework anchored in three major categories of possible unconstitutional constitutional amendment in Canada: procedural, substantive and procedural-substantive hybridity. Each of these three categories consists of at least three subsidiary forms of unconstitutionality. Procedural unconstitutionality includes subject-rule mismatch, temporal violations and processual irregularity. Substantive unconstitutionality includes unwritten unamendability, text-based unamendability and the amendment-revision distinction. Forms of hybrid unconstitutionality include statutory unconstitutionality, the recognition of convention and unconstitutionality by implication. My objective in this article is to offer the Court, litigators, political actors and scholars a road map to evaluate and apply the concept of an unconstitutional constitutional amendment in Canada.
I. Constitutional Amendment in Constitutional Democracies
Amendment rules serve an important cluster of functions that no other constitutional device can. They authorize a transparent process for correcting faults that may reveal themselves over time. (10) Amendment rules moreover distinguish the constitution from ordinary law, (11) the former usually insulated from change by more exacting thresholds and procedures. (12) Amendment rules also offer a way to check the judicial interpretation, (13) act as a vehicle to cultivate public discourse about constitutional meaning (14) and offer means to foster institutional dialogue among the branches of government. (15) Amendment rules may also be designed to express a hierarchy of constitutional rules reflecting a constitutional democracy's self-understanding of its public values. (16) Where they are especially onerous, amendment rules fill a pre-commitment function by making it difficult for political actors to change the entrenched preferences of the authoring generation. (17) At their core, formal amendment rules reflect the democratic values of the rule of law, providing notice and predictability to political actors and the relevant publics about who may change the state's most important political commitments, how they must do so and under what conditions. (18)
A. The Amendment Power in Constitutional Design
There is great variety in the design of formal amendment rules in constitutional democracies. (19) Formal amendment rules must necessarily differ according both to the particularized challenges confronting a constitutional state and to the stage of constitutional development in which the state finds itself, whether at its founding, on its way to democratic consolidation or as a mature democracy. (20) We may nevertheless observe, at a high level of abstraction, some important similarities among amendment rules in democratic constitutions. (21) Amendment rules are anchored either explicitly or implicitly in the foundational distinction between constitutional amendment and revision. They operate according to one of six frameworks that combine either single or multiple tracks of amendment procedures with restricted, comprehensive or exceptional rules about their use. They also entrench a combination of specifications such as voting thresholds, temporal limitations, electoral preconditions and subject matter restrictions.
(i) Between Flexibility and Permanence
The concept of constitutional amendment originated in the United States. (22) Early state charters and constitutions were the first to confront the possibility of their own imperfection. Amendment rules in the United States were created to give political actors a predictable and transparent method to make changes to these foundational texts. (23) America's first national constitution, the Articles of Confederation, entrenched a particularly difficult amendment rule requiring the approval of the national legislature and the unanimous agreement of all thirteen states. (24) This unanimity rule was perceived as a significant barrier to constitutional amendment, (25) and the veto it afforded each state in fact ultimately proved unworkable. (26)
When the delegates to the Philadelphia Convention gathered in 1787 to revise the Articles of Confederation, they had two related objectives in mind for the new amendment rule: first, to strike a federalist balance between the centre and the states, (27) and second, to promote constitutional durability by making the Constitution neither too easy nor too difficult to amend. (28) The result was Article V of the United States Constitution, which entrenches two major methods of formal amendment. (29) Under the traditional method, two-thirds of each house of Congress may propose an amendment that becomes valid when ratified by three-quarters of the states in either legislatures or conventions, as directed by Congress. Under the as-yet unused convention-centric method, two thirds of the states may petition Congress to call a constitutional convention in order to propose an amendment that becomes valid when ratified by three quarters of the states in either legislatures or conventions as directed by Congress.
(ii) Formal Amendment Rules and Modern Constitutional Democracy
Since the entrenchment of Article V in the United States Constitution, it has become common for national master-text constitutions to entrench formal amendment rules of their own. (30) Formal amendment rules influence constitutional politics even where political actors have no resort to them. (31) Constitutional rigidity may shift constitutional change from formal to informal mechanisms, pushing constitutional change "off the books". (32) Heather Gerken understands the relationship between formal and informal change in terms of hydraulic pressure: A rigid constitutional text that is not formally amendable "effectively redirects those constitutional energies into different, potentially more productive channels". (33) Those alternative channels include informal constitutional changes that result from quasi-constitutional statutes, treaties and constitutional conventions. (34) Amendment difficulty may also force constitutional courts to update or effectively "amend" the formally rigid constitution by interpretation, another species of informal constitutional change.
B. Formal Prohibitions on Constitutional Amendment
The power of formal amendment is rarely unlimited. Constitutional states commonly entrench prohibitions on the objects and subjects of the formal amendment power. For example, the French Constitution prohibits amendments to republicanism and to the integrity of the national territory. (35) Similarly, the Brazilian Constitution forbids amendments abolishing federalism. (36) The German Basic Law entrenches the best known example of a formal amendment prohibition, barring amendments that violate human dignity. (37)
(i) Designing Formal Unamendability
There are many reasons why constitutional designers might choose to make a constitutional provision impervious to the textually entrenched rules for formal amendment, even where there is overwhelming support from political actors and the public to amend it. First, they may wish to impose a gag rule on a particularly contentious matter, freezing the terms of agreement so as to free the parties to negotiate other parts of the constitutional bargain. (38) One example is the temporarily unamendable slave trade clauses in the United States Constitution, (39) negotiated in 1787 as a temporary resolution to a divisive matter. (40) Second, making something unamendable is a way for constitutional designers to entrench and thereby express to the world the constitutional values they believe do or should reflect the core identity of the constitutional state. (41)
Unamendability may serve three additional purposes: to preserve something distinctive about the state, to transform the state and to promote or accelerate reconciliation. (42) Constitutional designers may use unamendability to preserve what they view as an integral feature of the state, for example Islamic republicanism in Afghanistan. (43) They may also use it to transform the state, for example to repudiate an old regime and entrench a new political commitment, as the Constitution of Bosnia and Herzegovina sought to do by making all human rights formally unamendable. (44) Constitutional designers may also grant unamendable protections of amnesty or immunity for prior conduct as a way of encouraging reconciliation. The former Nigerien Constitution, for example, gave unamendable grants of amnesty to perpetrators of previous coups. (45)
Of course, no constitutional provision is really ever unamendable. Where the political will exists to alter an obdurate constitutional text, political actors can write a new constitution with the unamendable provision removed or loosened. This would break legal continuity in the regime, but it would nonetheless overcome the rigidity of the constitutional text. And where constitutional replacement is either impossible, improbable or suboptimal, the authoritative arbiter of constitutional meaning may stretch the interpretation of a constitutional law or amendment, finding as a result of creative interpretation that it respects the formal prohibitions set by the constitution, even if a plain reading would otherwise raise a tension with the formal prohibition. (46) This is precisely what occurred recently in Honduras: The Supreme Court interpreted as freely amendable a textually airtight formally unamendable clause prohibiting presidential re-election. (47)
(ii) Interpreting Formal Unamendability
The power to evaluate the constitutionality of a constitutional amendment may in theory rest with any political institution, but it often belongs to courts, and less commonly to legislatures. (48) Some unamendable provisions are more definitive than others and, as a consequence, leave comparatively little room for interpretation. Consider, for example, the Algerian Constitution, which makes the national language unamendable, (49) a rule that is more straightforward to interpret than the Namibian Constitution's absolute prohibition on any amendment that "diminishes or detracts" from fundamental rights. (50)
There are two major categories of violations of formal unamendability: procedural and substantive unconstitutionality. The Turkish Constitution, for example, entrenches secularism against formal amendment. (51) It also authorizes the Constitutional Court to review the constitutionality of amendments, but it limits the Court's review to only matters of form, that is to whether the amendment was adopted using the correct procedures, with the proper majorities and in the right sequence without irregularity. (52)
The Turkish Constitutional Court has been criticized for venturing beyond this pure procedural review of formal amendments, despite what appears to be a clear prohibition against a broader review of the substance of secularism. (53) Yet the distinction between procedure and substance is less clear than it appears since procedural rules often reflect substantive restrictions, and indeed there may be substantive values underpinning the procedures themselves. (54) Moreover, the Constitutional Court's review of substantive constitutionality despite its textual command to review only procedural constitutionality may reflect the difficulty of interpreting an unamendable provision like Turkey's unamendable value of secularism. Although secularism may be constitutive of Turkish constitutional identity, its meaning may, over time, vary even as its text remains unchanged.
In contrast to procedural review, a court may also review laws and constitutional amendments for substantive conformity with the constitution's formally unamendable rules. Germany, for example, makes human dignity inviolable and thereby creates a standard against which all official conduct, including laws and constitutional amendments, must be judged. (55) The German Constitutional Court has held that human dignity requires the state to protect pre-natal life over the mother's autonomy interest (56) and has relied on this constitutional value of human dignity to invalidate several others laws. (57)
A court may also interpret formal unamendability in connection with the adoption of a new constitution. The most well-known example comes from South Africa. In the transitional period after the end of apartheid, political actors adopted an interim constitution on the understanding that a new constitution would be adopted within two years of the first sitting of the National Assembly. (58) The interim constitution itself required the Constitutional Court to ensure that the eventual new constitution comply with a list of over thirty constitutional principles identified in an accompanying schedule. (59) The interim constitution made the Constitutional Court's certification decision "final and binding". (60) When the Court ultimately reviewed the new constitution for compliance with these constitutional principles--including the protection of fundamental rights, the separation of powers, judicial independence, federalism and the rule of law--the Court held that roughly ten items had failed to meet the standard. (61) The result was momentous: The new proposed constitution was unconstitutional. (62)
C. Informal Restrictions on Constitutional Amendment
Constitutional democracies sometimes recognize unamendability even where it is not entrenched in the constitutional text. In these cases, unamendability becomes informally entrenched as a result of a binding declaration by the authoritative interpreter of the constitution that something in the constitution is inviolable. The interpreter thereafter acquires the power to invalidate any contrary action, law or formal amendment. These informal restrictions rest on the fusion of two roles that have traditionally been separated across time and institutions: constitutional author and constitutional interpreter.
(i) Amendment and Revision
The distinction between amendment and revision is critical for understanding how informal unamendability arises. Carl Schmitt explained that an amendment occurs "only under the presupposition that the identity and continuity of the constitution as an entirety is preserved". (63) An amendment may be either ordinary or extraordinary in its effect, provided that in either case it remains continuous with the existing constitution and does not offend "the spirit or the principles" of the constitution. (64) An amendment, then, may expand, retract, specify or generalize as along as it "preserve[s] the constitution itself". (65)
Where a constitutional change alters the identity of the constitution, or runs counter to its spirit or principles such that the change transforms the existing constitution, that change is properly defined as a revision. A revision breaks with the fundamental presuppositions of the constitution and fails to cohere with its operational framework. (66) To illustrate, as John Rawls argued, it would be a revision to the United States Constitution, not an amendment, to repeal the First Amendment using the formal procedures of constitutional amendment. (67) The First Amendment, he suggested, should be understood as implicitly unamendable because it forms the core of the democratic presuppositions of the Constitution. (68) Of course, nothing in Article V prevents political actors from using its procedures to pass a hypothetical Twenty-Eighth Amendment doing away with the First. But the theory of revision regards the normal procedures of amendment as insufficient to authorize the eradication of a right so central to American constitutionalism.
Where the constitutional text does not expressly distinguish between procedures for amendment and revision, a judgment must be made whether a proposed change qualifies as an authorized amendment or whether it amounts to a revision. If it is determined to be an amendment, it is likely to become entrenched in the constitution without sustainable objection. But if it is determined that the change amounts to a revision, and that political actors attempted to revise the constitution using the procedures designed for amendment, the revision is likely to be contested. The basis for this informal restriction on the amendment power is the theory that a provision can be unamendable even where an amendment to it is not expressly prohibited in the codified constitution. This effectively creates an unwritten analogue to formal unamendability.
(ii) The Basic Structure Doctrine
The Supreme Court of India is best associated with the concept of informal unamendability. The Court refined the idea in a series of important judgments from 1951 to 1981. (69) Faced with the threat of the legislature abusing its textually unlimited power of formal amendment, the Court ultimately ruled that the amendment power was limited. The Court created the "basic structure doctrine" to invalidate amendments that, in its view, are inconsistent with the Constitution's framework. The Chief Justice wrote that "every provision of the Constitution can be amended provided in the result the basic foundation and structure of the Constitution remains the same". (70) What is most important to note is that the Constitution's text did not then, nor does it now, identify what is "basic", as in foundational, to its own structure. (71) That judgment of constitutional priority finds its origin and expression in judicial interpretation, not in popular consent-driven constitutional design.
II. Constitutional Amendment in Canada
Canada's formal amendment rules are among the most complex in the democratic world. (72) Their escalating thresholds, quorum requirements, opt-out exemptions and special protections for certain rights, institutions, structures and principles create a unique framework for amendment. Yet Canada's formal amendment rules stand out as much for what they entrench as what they do not: Unlike over half of the world's new recent constitutions, Canada does not entrench any form of formal unamendability. (73) Formally amending the Constitution of Canada is nevertheless no small feat, (74) and indeed it may be the most difficult democratic constitution to formally amend, (75) harder even than the United States Constitution, widely thought to be the world's most rigid. (76)
A. Formal Amendment in Canada
By its textual imprint alone, Canada's formal amendment rules are unique. The Constitution Act, 1982 contains sixty-one sections divided into seven parts. The fifth part, covering twelve sections and representing one-fifth of the entire text, is devoted exclusively to the rules for formal amendment. (77) It is unusual for democratic constitutions to entrench formal amendment rules in such length. The world's longest-enduring democratic constitutions entrench much shorter formal amendment rules, often in one or two sections. (78) This pattern generally holds in more recently adopted constitutions in democratic states. (79) Yet what distinguishes the Constitution of Canada's formal amendment rules from others in the modern democratic world is its combination of tiered voting thresholds, strict quorum requirements, opt-out exemptions and special protections for certain rights, institutions, structures and principles. These unique features are reflected in the escalating, federalist and consultative structure of formal amendment in Canada.
(i) The Escalating Structure of Formal Amendment
Escalation is the defining feature of Canada's formal amendment rules. The text formalizes five amendment procedures, each one expressly restricted for amendments to specific constitutional provisions and principles. This is an important feature of Canada's amendment rules: They do not have comprehensive application in the way we would describe the application of the amendment procedures in the Italian Constitution, for example, whose multiple formal amendment procedures may each be used to amend any formally amendable constitutional provision in the entire constitutional text. (80) Of the five formal amendment procedures in Canada, one applies exclusively to provincial constitutions: The legislature of each province is authorized to formally amend its own constitution in relation only to purely provincial subjects. (81) Each of the other four amendment procedures is more onerous than the other, and each is by and large cumulative in that it incorporates the requirements of the lesser one. (82) This is what I mean by escalation.
The lowest amendment threshold is the unilateral provincial amendment procedure in section 45. (83) The next-lowest amendment threshold is the unilateral federal amendment procedure in section 44. Under this procedure, Parliament is authorized to formally amend the Constitution by ordinary legislation "in relation to the executive government of Canada or the Senate and House of Commons". (84) Only the House of Commons or the Senate may initiate an amendment under this procedure, and both houses must approve the amendment before it receives Royal Assent. (85) In my view, Ian Greene has interpreted this procedure correctly: Parliament may deploy this amendment procedure to formally amend matters within its internal constitution, for instance Parliamentary privilege, legislative procedure and the number of members of parliament. (86) It is therefore an exceptionally narrow power--a limited delegation of power to Parliament--because it requires relatively little breadth of political support in order to be used successfully. (87) Further evidence of its thinness is evident in the Constitution's text, which makes its use "subject to sections 41 and 42", both reserved for more substantial formal amendments. (88)
The next amendment procedure in terms of difficulty is the regional amendment procedure in section 43, which incorporates the major elements of the unilateral federal amendment procedure. This Parliamentary-provincial procedure requires the House of Commons, Senate and legislative assemblies of the affected provinces to approve all amendments that apply to "one or more, but not all, provinces". (89) This procedure is more onerous than the unilateral federal procedure because it requires provincial consent, not just simple approval by both houses of Parliament. It must be used for amendments in relation to provisions of the Constitution that apply to one or more, but not all, provinces. Thus, it applies to matters that have, at a minimum, a provincial-federal interest even if it is in respect of a single province and, at most, to matters that have a regional, though not national scope. This procedure has been used more frequently and successfully than any of the four other procedures since 1982. (90)
The next most difficult amendment procedure is Canada's default amendment procedure. It applies to all subjects not otherwise assigned to a specific amendment procedure. It also applies exclusively to a specially designated class of subjects, including proportional representation in the House of Commons, the powers and membership of the Senate (as well as the method of senatorial selection), the Supreme Court of Canada for all items except its composition, the creation of new provinces and the boundaries between provinces and territories. (91) This default procedure in section 38 requires multilateral approval from both federal and provincial institutions: authorizing resolutions from the House of Commons and the Senate as well as resolutions from the provincial legislative assemblies of at least two-thirds of the provinces whose aggregate population amounts to at least half of the total. (92) This default amendment procedure incorporates the regional amendment procedure--though it does not give a veto to all provinces to which the amendment would apply, unlike section 43. It also requires supermajority provincial ratification as well as the majority population quorum requirement. (93)
Unanimity is the most difficult formal amendment threshold. Entrenched in section 41, it requires authorizing resolutions from both houses of the federal Parliament and from each of the provincial legislative assemblies. (94) There are five specifically designated subjects for which use of this unanimity amendment rule is required: the monarchy, the right to provincial representation in the House of Commons not less than that in the Senate, Canada's official languages beyond their provincial or regional use, the composition of the Supreme Court of Canada and Canada's formal amendment rules themselves. (95) This unanimity threshold is even more demanding than the default multilateral amendment procedure in section 38 insofar as it requires the approval of both houses of the federal Parliament and all ten provincial legislative assemblies. The Senate's approval is not an absolute requirement here. (96) We can therefore appropriately describe Canada's structure of formal amendment as escalating: Each of the four federal procedures requires more than the former, on the theory that the more important or politically salient a subject, the greater the degree of publicly aggregated political support required for making changes to it. (97)
(ii) The Federalist Structure of Formal Amendment
Canada's commitment to federalism is reflected in its escalating structure of formal amendment. What makes formal amendment more difficult as the subject or object of amendment rises in importance is the degree of provincial consent required to ratify an amendment proposal. After the unilateral provincial amendment procedure, the lowest of the other four thresholds is the unilateral federal amendment procedure which requires no direct provincial consent, though provincial interests are represented indirectly through the parliamentarians who vote on the amendment. The regional amendment procedure introduces the requirement of provincial consent but only for the affected province(s), that is, the province(s) to which the amendment applies. The quantum of provincial consent rises for the default multilateral amendment procedure, and of course rises to its highest point in the unanimity amendment procedure. This escalating structure of formal amendment is only one of many federalist features of Canada's formal amendment rules. Others include the right to register provincial dissent, the power to opt out from successful amendments in certain circumstances and in some cases to receive compensation for opting out, and the right to revoke both provincial dissent and assent. A word on each federalist feature is useful.
First, Canada's formal amendment rules authorize a province to register its dissent from an amendment made under the default multilateral amendment procedure. (98) There are three important qualifications to the right to dissent: (1) a province may only dissent from an amendment that weakens provincial powers or prerogatives, specifically one that "derogates from the legislative powers, the proprietary rights or any other rights or privileges of the legislature or government of a province", (99) (2) the dissenting province must pass a resolution by majority vote in its legislative assembly approving the province's dissent prior to the proclamation of the amendment, (100) and (3) the right of dissent is ineffective against formal amendments to proportional provincial representation in the House of Commons, Senate powers and provincial representation, senator selection and eligibility, the Supreme Court of Canada, provincial-territorial boundary modification and the creation of new provinces. (101) The effect of a provincial dissent is to grant the dissenting province an exemption from the application of the amendment: An amendment from which a province dissents "shall not have effect" in the dissenting province. (102)
The right to register provincial dissent amounts to a provincial power to opt out from an amendment that will otherwise apply to the entire country. For some matters, where a province registers its dissent to an amendment passed pursuant to the default multilateral amendment procedure and therefore opts out of the effect of the amendment, a province may be entitled to compensatory funding from the federal government. The Constitution authorizes the disbursement of "reasonable compensation" for a dissenting province where the amendment transfers from provincial control to federal control certain powers concerning education or culture. (103) For example, political actors might agree, by amendment pursuant to the default multilateral amendment procedure, to transfer jurisdictional authority over education from provincial legislatures to Parliament. Were the province of Ontario to register its dissent to that amendment, Ontario would be entitled to public funding to continue operating its provincial school system while education in the rest of the country would be overseen by Parliament.
The right of revocation is an additional federalist feature of Canada's formal amendment rules. The Constitution preserves provincial autonomy by conferring upon a province the right to revoke either its assent or dissent to a formal amendment. Where a province exercises its right to dissent, it may later revoke its dissent and simultaneously or subsequently consent to the amendment with a resolution supported by a majority of its legislative assembly. (104) There is no time limitation on the right of revocation; a province may revoke its dissent either before or after the proclamation of an amendment. (105) Provinces are not alone in possessing the right of revocation. Either the House of Commons, the Senate or a provincial legislative assembly may revoke a prior resolution assenting to an amendment as long as the revocation occurs before the official proclamation of the amendment. (106)
But the Constitution disables the twin provincial rights of revocation and dissent for specially designated provisions requiring the default multilateral amendment procedure. (107) For example, the default multilateral amendment procedure must be used to amend the provincial distribution of senators. (108) A province cannot dissent to, and therefore opt out of, a formal amendment to the number of senators to which it or another province is entitled. (109) To allow a province to dissent from the distribution of Senate seats would disrupt the structure and operation of Parliament. The same problem arises with respect to the principle of proportionate provincial representation in the House of Commons, which may be amended only with the default multilateral amendment procedure. (110) For the same reasons a province cannot dissent or opt out from Senate seat distributions, a province cannot dissent or opt out from an amendment to the scheme of proportionate provincial representation in the lower house. (111) The Constitution prudently anticipates problems arising out of this power to dissent.
(iii) The Consultative Structure of Formal Amendment
Formal amendment in Canada is also consultative by design, requiring political actors to consult deliberatively and cooperatively on major constitutional change. The first consultative dimension of formal amendment applies exclusively to the default multilateral amendment procedure as a temporal limitation. For formal amendments made pursuant to the default multilateral amendment procedure, the governor general may not issue a proclamation before one year has elapsed from the adoption of the resolution initiating the formal amendment procedure. (112) The governor general is also prohibited from issuing a proclamation after three years have elapsed from the adoption of the initial authorizing resolution. (113) This creates a one-year floor and a three-year ceiling for deliberation, the consequence being that all amendments proposed under the default multilateral amendment procedure expire after three years. This temporal limitation encourages purposeful debate within a defined period of time--political actors can neither rush nor delay an amendment.
The second notable consultative feature authorizes amendment without Senate approval. Where the Senate has not adopted an approval resolution for a formal amendment within 180 days after the House of Commons has done so, and the House of Commons once again adopts the same approval resolution sometime after 180 days, the amendment process may proceed without Senate approval. (114) Senate approval is normally required in all federal amendment procedures. This exception applies to all formal amendments made pursuant to either the unanimity amendment procedure, the default multilateral amendment procedure, or the regional amendment procedure. (115) It wisely does not apply to the unilateral federal amendment procedure (116) because allowing a formal amendment to pass without Senate approval would confer an unchecked power of formal amendment by legislation upon the House of Commons. The 180-day override power possessed by the House of Commons is therefore a mechanism to overcome political obstruction or delay by the Senate. Though it may appear contrary to the function of consultation, it furthers it by ensuring that political actors actually do consult, and ultimately decide, within a reasonable period of time. It is a constitutionalized protection against deliberate or passive delay.
Two additional features reflect the consultative dimension of formal amendment in Canada, yet they are not entrenched in the formal amendment rules themselves. First, the Constitution Act, 1982 required the prime minister to convene a constitutional conference with first ministers within fifteen years of its coming into force in order to review the formal amendment rules. (117) The authors of the Constitution Act, (1982) therefore contemplated multilateral consultation on whether, after almost a generation in use, the new formal amendment rules were serving Canada well. (118) Second, the Constitution Act, 1982 commits the prime minister to convene a constitutional conference consisting of first ministers and "representatives of the Aboriginal peoples of Canada" before any amendment is made to matters affecting Aboriginal rights. (119)
B. Judicial Restrictions on Constitutional Amendment
The text of the Constitution of Canada is not the only source of rules governing constitutional amendment. The Supreme Court of Canada has suggested that it could evaluate the process of constitutional amendment to ensure that the correct procedure is being used for amendments to the appropriate principle or provision. The Court has also created its own rules that both supplement and refine the formal amendment rules entrenched in the constitutional text. In the course of its interpretation of the Constitution, the Court has declared that certain unwritten constitutional principles are fundamental and suggested that they must be addressed in any negotiations leading to constitutional amendment. In a recent advisory opinion, the Court entrenched itself against amendment by all but the most onerous procedure of amendment. In this section, I review each of these to show how amendment is now constrained beyond the text.
(i) Procedural Enforcement
The Court asserted its authority to declare which constitutional amendment procedure must be used to amend a given principle or provision in the Senate Reform Reference when invited by Parliament to address the issue. (120) This raises two separate questions: whether the Government may seek the Court's counsel on a matter of constitutional law (it may and often has) (121) and whether the Court may specify that one amendment procedure must be used over another in a given instance. The answer to the second question is yes, but the Constitution does not expressly delegate this authority to the Court. It has arisen partly as a consequence of the Court's reference jurisdiction, its supremacy in constitutional interpretation and the latent ambiguities in Canada's formal amendment rules.
The central question in the Senate Reform Reference was whether one or another amendment procedure should be used to affect a series of separate changes to the Senate of Canada. The content of the question differed as to each envisioned senatorial reform, but in the end, the Court was asked to answer which of the Constitution's five amendment procedures political actors were required to use in a given scenario. In each instance, the Court answered clearly which amendment rule the circumstances dictated. This was the first major constitutional controversy since the enactment of the Constitution Act, 1982 in which the Court wrestled with the details of the entrenched formal amendment rules.
For example, in the Senate Reform Reference, the Government of Canada asked the Court for its advice specifically on whether the Constitution could be formally amended to establish fixed senatorial terms, for instance terms of eight, nine or ten years--using the unilateral federal amendment procedure in section 44. (122) The Court answered no, (123) explaining that the unilateral federal procedure is "limited", and that it "is not a broad procedure that encompasses all constitutional changes to the Senate which are not expressly included within another procedure in Part V". (124) This was a direct response to the unsuccessful argument that Parliament could amend senatorial term limits unilaterally since there is no express mention of term limits in the rules of multilateral formal amendment pertaining to the Senate. Term limits are amendable, explained the Court, but only using the default multilateral amendment rule in section 38. (125) There is good reason for not allowing Parliament to use its limited power of formal amendment to change senatorial tenure: The change touches on a matter of federal-provincial interest, and therefore any process to alter it must engage both levels of government, not only Parliament.
The Court gave a similar answer to the question of whether Parliament could effectively change the method of choosing senators. (126) The government asked the Court whether Parliament could use the unilateral federal procedure to create a framework for consultative senatorial elections that would authorize the populations of the provinces and territories to express their preferences for senatorial nominees. (127) In answering no, the Court took a functionalist view of constitutional change, reasoning that a constitutional amendment need not necessarily alter the constitutional text in order to alter the constitution. It would "privilege form over substance", wrote the Court, to define an amendment so narrowly, even where the governor general would continue to appoint--in the language of the Constitution, to "summon"--senators on the recommendation of the prime minister. (128) The Court held that introducing these changes to the method of senatorial selection would so fundamentally alter the federal architecture of the Constitution that Parliament could not alone make this change. (129) The provinces, the Court explained, must be involved in this amendment process pursuant to the default multilateral amendment procedure in section 38. Indeed, the Court added, the text of the amendment rules requires that changes to the "method of selecting Senators" be made in consultation with the provinces, not unilaterally. (130)
The other reference questions likewise asked the Court which amendment procedure was proper for a given reform to the Senate, namely repealing the property qualifications for senators and abolishing the Senate altogether. (131) On the former, the Court decided that the Constitution allowed Parliament to use its unilateral amendment power, although for political actors to fully repeal the Senate they would need also to use the regional amendment procedure to secure the consent of Quebec, which has a special arrangement in respect of the Senate. (132) On the latter, the Court explained that abolishing the Senate would require conformity with the onerous rules of the unanimity amendment procedure, namely the consent of the houses of Parliament and all of the provinces. This is precisely because of the importance of the Senate to the structure of the Constitution, to federalism in Canada and to the design of Canada's formal amendment rules themselves. (133)
(ii) Unwritten Constitutional Principles
The Supreme Court has also identified certain unwritten constitutional principles that constitutional amendments must respect. This extraordinary move has incorporated unwritten rules into the written constitution, and arguably subordinated the text to them. The result has been to create a hierarchy of constitutional precedence placing unwritten rules above written rules, raising concerns for the rule of law--itself a constitutional principle. As Lon Fuller argued, the rule of law requires the publication of clear rules so the governed have notice of their obligations and entitlements. (134) Where unwritten principles are given priority over written rules, there is a risk of a disjunction between law and enforcement (135)--a disjunction that can undermine the rule of law. Whether this risk has materialized in Canada in the context of constitutional amendment remains an open question. What is less uncertain, however, is that unwritten rules are often uncovered from "amendment-like" interpretations of the Constitution. (136)
For example, in the Secession Reference, the Court identified four unwritten constitutional principles that must govern negotiations leading to a constitutional amendment to formalize secession. (137) By the Court's own admission, these constitutional principles--federalism, democracy, constitutionalism and the rule of law, and the protection of minorities (138)--appear nowhere in the express provisions of the constitutional text. (139) As the Court wrote, these principles "are not explicitly made part of the Constitution by any written provision", but it "would be impossible to conceive of our constitutional structure without them" because they "dictate major elements of the architecture of the Constitution itself and are as such its lifeblood". (140) In light of the evolution of Canadian constitutional history, (141) these principles give rise to a reciprocal obligation on all parties to negotiate a provincial secession when a clear majority of a province has chosen secession on a clear referendal question. (142) These four unwritten principles are not merely descriptive. They are so important that they may constitute a legitimate basis for invalidating the conduct of political actors.
Underlying constitutional principles may in certain circumstances give rise to substantive legal obligations, which constitute substantive limitations upon government action. These principles may give rise to very abstract and general obligations, or they may be more specific and precise in nature. The principles are not merely descriptive, but are also invested with a powerful normative force, and are binding upon both courts and governments. (143) The Court here recognized the existence of these unwritten constitutional norms, and also kept open the possibility of recognizing others, and in doing so has preserved for itself wide latitude to police all forms of state action. As the Court itself stressed in the Secession Reference, the Court is authorized to invoke these and other unwritten principles in all matters that arise before it. (144)
This was not the first time the Court had invoked unwritten principles as a decision rule. The year prior, in the Provincial Judges Reference, the Court seemed to suggest that a law could be invalidated for violating an unwritten constitutional principle, in this case judicial independence. (145) The Court insisted that although the Canadian Charter of Rights and Freedoms (146) and the Constitution Act, 1867 both give a limited guarantee of an "independent and impartial tribunal", (147) judicial independence "is at root an unwritten constitutional principle, in the sense that it is exterior to the particular sections of the Constitution Acts'". (148) Drawing from precedent and the nature of Canadian constitutionalism, the Court concluded that the Constitution includes more than its entrenched texts. The Court thus cautioned against presupposing that "the express provisions of the Constitution comprise an exhaustive and definitive code for the protection of judicial independence". (149) The preamble of the Constitution Act, 1867, whose purpose is partly "to fill out gaps in the express terms of the constitutional scheme", (150) also reinforces judicial independence as an important norm.
Rooted in history, entrenched in the text and anchored in the preamble of the Constitution Act, 1867, judicial independence was not the only unwritten constitutional principle the Court identified in this case. The preamble also identifies other organizing principles that have similar normative validity, according to the Court. For instance, the doctrine of "full faith and credit", obliging provincial courts to recognize the judgments of others, must be inferred from the Constitution and its preamble since it is fundamental to federalism in Canada though not expressly entrenched in the constitutional text. (151) Likewise, paramountcy--which holds that a valid federal law prevails over a valid provincial law to the extent of any inconsistency--appears nowhere as a general proposition in the text, but it is a necessary feature of Canada's federalist constitutional design. (152) The Court followed similar reasoning to support its recognition of other unwritten constitutional norms, including the rule of law's remedial innovation of suspended declarations of invalidity, the constitutional status of the privileges of provincial legislatures, the federal power to regulate political speech, and implicit limits on legislative sovereignty with respect to political speech. (153) There are others still to be uncovered and applied, though without the same clarity that only a text can offer. (154)
(iii) Judicial Self-Entrenchment
In addition to, but quite apart from entrenching unwritten constitutional principles, the Court has also entrenched itself. Yet in self-entrenching, the Court has not made itself formally unamendable: The Court remains amendable by political actors, specifically by the default multilateral amendment procedure, and by the unanimity procedure as to its composition. (155) Nonetheless, in the recent Supreme Court Act Reference, the Court made it much more difficult than the text of the Constitution suggests it should be to make amendments to the Court itself. By its interpretation of formal amendment rules, the Court has in this way entrenched itself against ordinary amendment. (156)
Reference re Supreme Court Act, ss 5 and 6 concerned two inquiries related to eligibility for a seat on the Supreme Court. The first was whether a person who was currently a judge on the Federal Court of Appeal but had previously been, though was not currently, a qualified Quebec attorney for at least ten years could qualify as a Quebec judge under the Supreme Court Act, which reserves three seats for judges appointed "from among the judges of the Court of Appeal or of the Superior Court of the Province of Quebec or from among the advocates of that Province". (157) The Court answered no, concluding that the law's requirement that the appointee be a qualified Quebec attorney means that the appointee has to be a current member of the Quebec bar, at the time of the appointment, with at least ten years standing. (158) The second question, though, is more relevant for our purposes. The Court was asked to advise the government whether in light of the answer to the first question, Parliament may pass a law remedying that ineligibility, thereby authorizing the appointment of a former member of the Quebec bar to the Court. (159) The Court again answered no. (160)
In detailing why Parliament cannot authorize by simple law an appointment not otherwise permitted by the Supreme Court Act, the Court had its Canadian Marbury moment--a reference to Marbury v Madison, (161) the famous case in which the United States Supreme Court declared itself not only the authoritative interpreter of the constitution, but also the ultimate arbiter of its own jurisdiction. Just as Marbury illustrates an example of self-entrenchment, so does the Supreme Court Act Reference. The Canadian Supreme Court positioned itself as the only body that can constitutionally interpret the Constitution as to others and as to itself, a position that could conceivably raise a conflict, but that one could justify under section 52 of the Constitution Act, 1982, which makes the Constitution of Canada supreme (162) and by implication the Court's interpretation of it as well. What the Court ruled in the Supreme Court Act Reference can be stated quite plainly as follows: Parliament cannot by itself repeal or even amend the Supreme Court Act, except for routine amendments that do not affect the essential characteristics of the Court.
That Parliament cannot alone amend its own law would once have been controversial and thought to uproot the very foundations of Parliamentary sovereignty that Canada inherited from the United Kingdom. But Parliamentary sovereignty in Canada has never implied Parliamentary supremacy given the jurisdictional limitations imposed on Parliament and the provincial legislatures by the Constitution Act, 1867. Since 1982, Canada has been a constitutional supremacy, with the consequence that Parliament and the provincial legislatures must now conform their conduct to the Charter, which has imposed new obligations on parliamentarians. (163) One of those obligations is to abide by the Court's interpretation of the Constitution, which has long included more than its text. Since the Supreme Court Act Reference, however, the Constitution arguably now includes the Supreme Court Act itself. That is the operational result of the Court's opinion in the Supreme Court Act Reference: It elevates most aspects of the Supreme Court Act beyond ordinary Parliamentary action by informally entrenching the key characteristics of the Court. Today, amending the Court's essential features requires a multilateral constitutional amendment.
The composition of the Supreme Court, wrote the majority, cannot be subject to simple Parliamentary legislative amendment. (164) Any change to the Court's composition, which for the Court included a change to the rules for the three Quebec appointments, must be made using the unanimity amendment procedure in section 41. (165) Otherwise, Parliament could unilaterally amend the essential features of the Court and thereby risk undermining the Court's independence, its function in the separation of powers and its power as the authoritative interpreter of the Constitution. As the Court explained, "essential features of the Court are constitutionally protected" and any changes to the Court's composition require "the unanimous consent of Parliament and the provincial legislatures". (166)
The practical consequence of the Court's opinion is significant. The Court has effectively transformed a Parliamentary law into a constitutional statute that now forms part of the Constitution of Canada, (167) causing the "essential features of the Court" to "migrate" into the Constitution, where it is now immune from anything less than multilateral constitutional change. (168) The Court has recognized that the evolution of Canadian federalism and of the Court's role within it as the national court of last resort now requires that its composition and its fundamental components be protected from ordinary Parliamentary law making. (169) Parliament, the Court ruled, cannot have the power to make transformative changes unilaterally, either to the Court as an institution of central importance to Canadian federalism or to Quebec's historically guaranteed representation. (170) Conferring this power upon Parliament, the Court concluded, would ignore the Court's constitutional status and modern constitutional politics. Moreover, it would deny the Court the capacity to exercise its function under the Constitution. (171)
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Abstract through II. Constitutional Amendment in Canada B. Judicial Restrictions on Constitutional Amendment, p. 143-176; Constitutional Renewal: Comparative Lessons for Canada|
|Publication:||Queen's Law Journal|
|Date:||Sep 22, 2015|
|Previous Article:||Deliberative constitutional amendments.|
|Next Article:||The theory and doctrine of unconstitutional constitutional amendment in Canada.|