Peter C. Oliver: Constitutional Conventions in the Canadian Courts

Most constitutional law textbooks across the Commonwealth include in the section on constitutional conventions lengthy extracts from the Canadian case, Re Amendment of the Constitution of Canada, often referred to as the Patriation Reference.  Given that constitutional conventions are enforced not by courts but by the political process, conventions are rarely discussed in the law reports.  Courts occasionally recognize constitutional conventions in order to discuss intelligently the way the contemporary system of democratic responsible government works, but they do not accede to a party’s attempt to win a court battle by requesting that the court enforce a convention. Even in the well-known case of Jonathan Cape ((1975) 3 All ER 484, [1976] QB 752), the Attorney General would not have made any progress if he had only argued the convention (of collective ministerial responsibility); the success of his argument, so far as it went, was founded on the equitable doctrine regarding breach of confidence, with the convention supporting the factual finding of a confidential circumstance.

What was different about the Patriation Reference?  And what effect has that case had on the way Canadian courts view the justiciability of constitutional conventions? Readers of this blog may be aware that Canada has just lived through a period of minority government in which the Prime Minister’s decisions to prorogue and dissolve Parliament were hotly debated.  Political actors who were critical of the Prime Minister’s actions sought out any and all means thwarting him.  An organization called Democracy Watch, for example, challenged the 2008 dissolution of Parliament in court on the basis of both law (fixed election legislation) and convention.  Later in 2008, critics called for a Supreme Court of Canada advisory opinion regarding the Prime Minister’s controversial decision to avoid a vote of non-confidence by asking the Governor General to prorogue Parliament.

Canadian courts have so far resisted these and other calls for them to enforce constitutional conventions.  However, their resolve is not as clear and firm as it could be.  Since 2000, both the Supreme Court of Canada and the Federal Court of Canada have produced judgments (in the context of ordinary litigation rather than in an advisory role) in which the existence or non-existence of conventions is discussed in some detail, including analysis of whether Jennings’ three-part test (first adopted in the Patriation Reference) has been met. One cannot help but wonder what will happen if, as seems inevitable, a constitutional convention is eventually made out in the context of future litigation. Will the court say that it is merely recognizing the convention not enforcing it, and that, accordingly, the remedy is by way of declaration not injunction?  That would seem to misunderstand the way in which law is enforced in this day and age, especially where the government is a party: a mere declaration of the law is invariably all that is required.  Or will the court re-discover the essentially political nature of the question and declare it non-justiciable?  To do so, it would have to revisit not only the Patriation Reference which first opened up conventions to judicial consideration, but also the Secession Reference and similar Supreme Court of Canada cases which have opened up the use of principles in the fashioning what is sometimes referred to as the common law Constitution. After all, all conventions are underpinned by a principle (e.g. the democratic principle) according to Jennings’ three-part test cited in the Patriation Reference (at p. 888). What is to stop the principle, and the principle-inspired jurisprudence of the Supreme Court of Canada, from being used to transform convention into law?  How did we get from the Patriation Reference to here?

Prior to the Patriation Reference, it was as rare in Canada as it was and is elsewhere in the Commonwealth to see conventions discussed in judgments.  A number of factors help explain what happened in 1981. First, as is well known, Canada, unlike countries (e.g., U.S.A., Australia) which hold to a stricter version of the separation of powers, permits advisory opinions (or references).  The Canadian Supreme Court Act sets out in the widest terms imaginable (see s. 53) the sorts of questions that can be put before the Court. The constitutionality of these provisions was confirmed in the Secession Reference. Secondly, although the Supreme Court Act provides few hints in this direction, the Court reserves a discretion to refuse to answer questions, often on the basis of non-justiciability.  Justiciability is a question of the courts’ proper role, but it is also a question of the courts’ relevant expertise.  As it happens, any judge who wishes to understand a constitution in the British tradition must understand constitutional conventions.  As the majority on the conventional question put it in the Patriation Reference (pp. 883-4): Constitutional Law = the conventions of the constitution + the law of the constitution.  While determinations of the precise present state of a convention might be beyond most judges’ ken, most judges could be said to have considerable knowledge in this area nonetheless.  A conclusion of non-justiciability would perhaps have to come from the Court’s sense of its proper role vis-à-vis other institutions rather than from considerations of relevant expertise. Thirdly and finally, the stakes in 1981 were very high: Canada had struggled for fifty years to come to agreement on a new amending formula to replace the United Kingdom Parliament as its ultimate constitutional amendment procedure. A first referendum on Quebec’s future in the federation had just been held one year earlier.  And clearly a majority of judges of Supreme Court of Canada judges felt that they had to weigh in.

The problem with courts weighing in regarding constitutional conventions was already apparent in 1981.  When the Court determined that, as a matter of the law of the constitution, United Kingdom Parliament legislation amending the Canadian constitution was legally unobjectionable, that was in principle sufficient to justify Prime Minister Trudeau’s attempt to patriate the constitution over the objections of eight out of ten provinces.  However, when the court also stated that, as a matter of the conventions of the constitution, a substantial degree (p. 905) of provincial consent was required, the distinction between law and convention quickly faded. To all but the most informed, formalistic observers, the Court had spoken authoritatively, and it had to be followed, sending Trudeau and the provinces back to the negotiation rooms (from which they emerged, Quebec dissenting, with the compromise that became the Canada Act 1982, the Constitution Act, 1982 and the Canadian Charter of Rights and Freedoms.)

Staying for the moment with the Patriation Reference, we might want to note the way in which the judges who signed the majority reasons regarding the conventional side of that matter proceeded. They clearly felt that they had to answer the question, for the reasons that I have already set out and no doubt for other reasons.  However, the provinces opposing what they viewed as federal unilateralism asked for more than the Court was willing to give.  The provinces argued that constitutional amendments affecting provincial interests required the consent of all provinces, that is, unanimous consent. The majority may have felt that a convention pitched at such a high level was not clearly established.  It opted instead to pitch the convention at a level which could be clearly established, that is, a substantial degree of consent (which all observers assumed to be more than the existing two-province consent but less than unanimous consent).  Was the Court effectively saying that only a convention of at least substantial consent could be clearly established, but that there was an ongoing, evolving political discussion as to whether more was required, including unanimous consent?  If the Court wished to say this, it did not communicate it clearly, with the result that most observers assumed that the Supreme Court of Canada’s finding regarding convention was authoritative in the exclusionary or pre-emptive sense which Joseph Raz describes regarding law proper: the Court’s pronouncement effectively replaced the many discussions in the political arena regarding the political (i.e., conventional) morality of the pre-1982 amendment process.

Even though, as a rule in Canada, the law as stated in advisory opinions is followed by courts in ordinary litigation, there was no initial spate of litigation in which parties sought to vindicate in court their particular take on a constitutional convention.  After all, even as the Supreme Court of Canada had set out a constitutional convention in the advisory Patriation Reference, it had reiterated the orthodox rule regarding the fact that conventions are not enforced by courts (p. 880).  And the Court has never disavowed the orthodox rule.

What has changed since 1981, however, is the Court’s general jurisprudential approach.  When the Court spoke in 1981 of the law of the constitution, for example, it spoke from an essentially positivist perspective.  It was referring in the main to legislation, British and Canadian, and case law interpreting that legislation. By the mid-1990s, Canadian courts had adopted a principle-inspired method based, in part, it would seem, on the anti-positivist theories of Ronald Dworkin.  Where the law of the constitution leaves a gap (or, as Dworkin might say, presents a hard case), the Canadian courts are now less shy to fill the gap with the principle-inspired common law of the constitution. This begs the question whether the principles which underlie each and every constitutional convention can now be used to facilitate the gradual conversion of the conventions of the constitution into the law of the constitution.

I argue that the Canadian courts should ensure that this does not happen.  And yet if a principle-inspired legal method provides the tools to convert conventions into law why not?  In my view, the answer can be found in the preamble to the Constitution Act, 1867, which states that Canada has a Constitution “similar in principle to that of the United Kingdom”.  That phrase in the preamble signaled that while some parts of the Constitution were as of 1867 written, equally important elements remained mostly unwritten.  Or put another way, that while some parts of the Constitution were designed to be judicially enforced, other important elements were designed to be politically enforced.  With regard to the latter, one thinks of constitutional conventions, including those establishing responsible government, parliamentary privilege, and the very idea of parliamentary sovereignty.  The Canadian constitutional system has evolved, of course – it is a “living tree” – but the political Constitution is still an important part of our democratic constitutional set-up. The Constitution Act, 1982 replaced the pre-1982 conventions regarding constitutional amendment and limited parliamentary sovereignty, but it did not eliminate the important role of conventions, responsible government, privilege and parliamentary supremacy and the political Constitution as a whole.

Accordingly, the main reason why the courts should not use the principles underlying conventions to convert conventions into law is that there is no gap to fill.  That which appears to be a hard case is only so if one focuses exclusively on the law of the constitution.  If one considers together the conventions and the law of the constitution, the political and the legal constitution, then it is clear that judges should respect political means of enforcement rather than use the principles underlying conventions as a means of converting rules of political enforcement into rules of judicial enforcement.

Part of the difficulty here lies perhaps in lawyers’ general preference for text over practice.  Conventional rules, and much of the political constitution, look like a gap or an empty space from the perspective of the text-based legal constitution.  However, if we imagine a case where conventions are written down for ease of reference, but where it is clearly intended that enforcement remain political rather than legal, the courts would view the interpretation and enforcement of those rules as non-justiciable rather than view them as a gap or a hard case requiring conversion into new legal rules of the common law constitution by means of principle-inspired jurisprudential method.  The same approach should apply even where the conventions are not written down, until such time as a decision is made to convert conventions into law (as occurred regarding the conventions of constitutional amendment in Canada).

An additional source of concern, from my perspective, is that the Supreme Court of Canada has, with respect, turned the meaning of “a Constitution similar in principle to that of the United Kingdom” on its head.  I have already stated that, in my view (and in the view of constitutional historians and the Supreme Court of Canada itself prior to 1982) that phrase was intended as a reminder of the ongoing existence and importance of the political part of the Constitution.  In a number of important cases in the 1990s (see, e.g., the Provincial Judges Reference, para.104 and the Secession Reference, para. 53), the Court used the same phrase in the preamble as an “invitation” to the courts to fill perceived gaps in the Constitution using principles such as constitutionalism and the rule of law, democracy, federalism, the protection of minorities and judicial independence.  In fact, the Court seemed intent on finding a textual basis for its use of unwritten principles.  Unfortunately, the preamble text it used was itself a reminder of the importance of political as opposed to judicial enforcement of constitutional rules.  This is not to say that the Court should not use principles to fashion its decisions – that is a debate for another day. Rather, the point that I am trying to make is that the Court should refrain from using the very principle that underpins a conventional rule, and the very phrase from the preamble of the Constitution that is intended to recognize the ongoing importance of political enforcement of these constitutional rules, to create new parts of the common law constitution.

I have noted that the Supreme Court of Canada has already, with respect, (mis)used the phrase “a Constitutional similar in principle to that of the United Kingdom” to justify a principle-based approach to developing the common law of the Canadian constitution. Is there any sign, however, that the Canadian courts wish to go further, that is, to convert conventions into legal rules using the same method?  The signs are mixed, in my view.  In two cases in 2000-1 (Public School Boards’ Association of Alberta v Alberta (Attorney General), [2000] 2 SCR 409, paras 30 and 38; Ontario English Catholic Teachers’ Association v Ontario (Attorney General), [2001] 1 SCR 470, paras 26 and  63 et seq), the Supreme Court of Canada allowed parties in ordinary litigation to state a constitutional question regarding, inter alia, a constitutional convention.  In other words the parties were attempting to win their case by a number of means, one of which could have involved a declaration involving a constitutional convention.

One might have expected the court to simply refuse to answer the question.  It is one thing to ask a court to recognize a convention as part of the factual context necessary to understanding the legal dispute, and quite another to try to win the case on the basis of a convention.  Instead, after a reminder of the fact that conventions are not enforced by courts, the Supreme Court of Canada then went on to discuss Jennings’ test for constitutional conventions first set out in the Patriation Reference.  In my opinion, this gives potential litigants the impression that, if the convention can be made out in a future case, the Court might be willing to issue a declaration to that effect, though clearly it would refuse to award an injunction.  As I stated earlier, if that impression is right, then there is cause for concern.

Since 2000-1, conventions have continued to rear their head from time to time in litigation.  Following the Supreme Court of Canada’s ambiguous lead, the lower courts have not always felt able to refuse to discuss the granting of a declaration regarding a constitutional convention.  (See, e.g., Pelletier v Canada (Attorney General), 2007 FC 342, para 6,  and Pelletier v Canada (Attorney General) 2008 FCA 1, para 18 et seq;  Conacher and Democracy Watch v Canada (Prime Minister), 2009 FC 920, paras 2, 10-15, 30-47, 65-72; Conacher and Democracy Watch v Canada (Prime Minister), 2010 FCA 131, paras 5-6, 12, )  Instead, they have often restated the non-enforceability rule, but then gone on to discuss, sometimes in considerable detail, whether the convention is made out (applying Jennings and the Patriation Reference).  It seems to me that the courts’ approach to constitutional conventions needs to be more clearly thought through in advance of the day when a declaration regarding a credible constitutional convention is sought in ordinary litigation.  It seems inevitable that if the courts signal that they are willing to grant declarations, parties which have the financial means to litigate will seek to lift their cases out of the ongoing and evolving discussion in political forums by seeking discussion-stopping victory in the courts.

Accordingly, I would recommend the adoption of the following guidelines at least in so far as treatment of conventions by courts in Canada is concerned.  I would be interested to hear how constitutional lawyers in other jurisdictions view these suggestions.

  1. In the context of ordinary litigation, courts should refuse parties’ requests, by way of declaration or other remedy, to rule on the existence or non-existence of a constitutional convention in order to determine the legal outcome of the case.
  2. In the context of ordinary litigation, in which (consistent with 1.) the dispute turns on something other than the existence or non-existence of a constitutional convention, courts may sometimes find it necessary in the course of their reasoning to describe (or recognize) a well-established constitutional convention in order to sensibly account for our constitutional arrangements.
  3. In the context of a reference case (advisory opinion), the court should accept requests to rule on the existence or non-existence of a constitutional convention, only to the extent that the conventional rule is, in the court’s view, clear and well-established.  If the convention itself is in flux, or if the question relates to a peripheral aspect of the convention which is likewise still in flux, then the courts should regard the question as non-justiciable, both because of the courts’ lack of expertise regarding evolving political dynamics, and because of the need to maintain a proper balance between the judicial and political parts of the constitution.
  4. In the context of a reference case (advisory opinion), the court should perform an important educative function in explaining that many constitutional conventions, and the political parts of our constitution in general, are in constant evolution, and that the public’s democratic responses to perceived breaches of convention are critical to that political process.

These comments and recommendations are offered on the thirtieth anniversary of the Patriation Reference and on the fortieth anniversary of the publication of Constitutional Theory by my supervisor and good friend, the late Dr Geoffrey Marshall.

Peter Oliver is a Professor of Law at the Faculty of Law, University of Ottawa and the author of The Constitution of Independence: The Development of Constitutional Theory in Australia, Canada and New Zealand (OUP, 2005).  He was formerly Professor of Law at King’s College London.  The author was Scholar in Residence in the Constitutional and Administrative Law Section, Public Law Sector, Justice Canada in 2005-6 during which time he completed a research contract looking into the meaning  of the 1867 preamble phrase “A Constitution similar in principle to that of the United Kingdom”.  In 2006-7 he was Special Advisor, Legal and Constitutional Affairs at the Intergovernmental Affairs Secretariat of the Privy Council of Canada, and he has continued to act in that capacity from time to time from 2007 to present.  He has not advised on the issues discussed in this piece. The views expressed here are the author’s own and should not be taken to represent the views of the Intergovernmental Affairs Secretariat or of the Government of Canada.