Bill C-16, An Act to amend the Canada Elections Act, received royal assent on May 3, 2007. It provided that a general election would be held on the third Monday in October in the fourth calendar year following polling day for the preceding general election, thereby setting up a system of fixed election dates. It provided for the first such election to be held on Monday, October 19, 2009. However, the amended Act also stipulated that the powers of the Governor General were to be unaffected, in particular the power to dissolve Parliament at her discretion. In 2008, the Prime Minister asked the Governor General to dissolve Parliament, and the first general election following the amendment of the Act was held on October 14, 2008. That prompted an outcry from a number of observers, some going so far as to suggest that the new law had been infringed. An application was made to the Federal Court challenging the government’s action. The Court’s judgment refusing the application was handed down on September 17, 2009. This article looks at the issues raised by the parties and the decision of the Court.
The decision of the Federal Court in Conacher v. Canada (Prime Minister)1 offers a glimpse from a judicial perspective of a feature of the constitution that generally escapes the purview of the courts, despite its crucial role in the functioning of our legal institutions: the Crown’s power to dissolve Parliament and precipitate a general election. Wrapped up with it is the Governor General’s obligation under most circumstances to exercise her powers on the advice of the government of the day, an obligation founded not in law but in convention and therefore not enforceable by the courts.
The application for judicial review sought a declaration that the calling of the election in October 2008 was contrary to the new section 56.1 of the Canada Elections Act, 2 which ostensibly provides for a regime of regular fixed-date elections. The grounds of the application involved the interpretation of the statutory language, but also led the court to a consideration of the nature of the royal prerogative and constitutional conventions. It became necessary for the Court to consider at some length its jurisdiction to hear and determine the issues raised. In particular, it addressed an argument that the Governor General’s decision was ultimately political in nature and that judicial scrutiny of such actions would upset the “separation of powers” between the executive and judicial branches of government. The Court accepted this argument, among others, and denied the application.
Prerogative and Statutory Powers
The prerogative of the Crown, in Canada as in the United Kingdom, comprises the residue of royal powers and privileges that have survived over centuries as parliaments gradually became ascendant. These powers form part of the common law and so fall outside the “written” constitution, made up of enactments such as the Constitution Acts, 1867 to 1982. Prerogative powers generally relate to core executive functions or matters of “high policy” that require wide discretion and political judgment. Among the most significant are the choice of ministers of the Crown, the conduct of intergovernmental relations, issues relating to national defence, the direction of parliamentary business and the summoning, prorogation and dissolution of Parliament itself. In Canada, the Sovereign’s prerogatives are formally delegated to the Governor General, currently in “letters patent” issued in 1947 by George VI on the advice of Prime Minister Mackenzie King.3 By virtue of the conventions of parliamentary government, the Governor General as a rule exercises her powers, both statutory and prerogative, on the advice of the Prime Minister and Cabinet.4
In very limited circumstances, a Governor General may act without or even contrary to ministerial advice. Convention dictates that the Prime Minister is the party leader who, for the time being, can command a majority of the members in the House of Commons on a vote of confidence. In a minority Parliament, the Prime Minister and his party can lose such a vote, in which case he can either resign or request a dissolution. Resignation leads to a transfer of power to the incumbent’s adversaries, so there is a natural tendency to seek an election and the opportunity to improve one’s party standings. The Governor General retains a “personal” prerogative which operates as a brake against this tendency. Following a loss of the House’s confidence, she may decline advice to dissolve Parliament and instead select another party leader who stands a chance of gaining that confidence. The clearest occasion for such refusal would be the defeat of a government’s first throne speech following a general election. On the other hand, refusal would be considered unusual after a minority Parliament had survived as long as two years.
The major source of executive power, in terms of volume and visibility, is today found in statutes and regulations. The extensive powers conferred by enacted law on the Governor in Council (the Governor General acting on the advice of the Cabinet) tend to be specific and detailed. The nature of prerogative powers makes them less amenable to codification in a rigid set of written rules. Although they confer a wide latitude on government, it is not unlimited and when exceeded is subject to judicial review. The distinction between statutory and prerogative powers becomes an issue when governmental action is challenged in the courts.
The jurisdiction of the Federal Court to review illegal action under the Federal Courts Act is limited mainly to the exercise of powers conferred “by or under an Act of Parliament.”5 On its face, this language excludes the Governor General’s power of dissolution, being an element of the Crown’s prerogative rather than a power conferred by statute. The Federal Court appeared to appreciate this distinction but did not expressly rely on it to deny the application. A complicating factor is section 50 of the Constitution Act, 1867, which sets the maximum lifespan of a parliament at five years while expressly leaving intact the Governor General’s power to dissolve it at any time.6 Whether that section turns the prerogative power of dissolution into a statutory one is debatable. Assuming that it does, however, the Constitution Act, 1867 is not an Act of the Parliament of Canada, and so any powers conferred by it do not technically fall within the ambit of the Federal Courts Act.
Executive Authority and Judicial Review
Even if the applicants could establish that the Federal Courts Act applied to this particular government action, they faced a more fundamental hurdle. While many executive or administrative decisions are open to review by the courts on various grounds, the power to dissolve Parliament and proclaim an election has always been thought to be fundamentally and exclusively a political decision. As such it is not “justiciable” but is within the exclusive discretion of the executive. At the time the 2008 election was called, a minority government had been in power for more than two and a half years and it had survived a number of no-confidence votes. In such circumstances, convention called for the Governor General to follow her Prime Minister’s advice. Moreover, neither law nor convention restricted the Prime Minister’s discretion to request a dissolution. Was there anything in the new section 56.1 that altered long-established usages and precedents so as enable a court to second-guess the Prime Minister’s advice or the dissolution itself?
In the hearing before the Court, the applicants centred their case on the “Prime Minister’s decision”. The case of Black v. Chrétien in the Ontario Court of Appeal7 was cited as authority that a prerogative power could embrace advice given to the Governor General as well as the decision based on that advice. In the Black case, the plaintiff had challenged steps taken by a former Prime Minister to oppose the creation of a peerage under the Crown’s “prerogative of honours”. The Ontario Court of Appeal had held that this prerogative was among those that are inappropriate for judicial review. It cited an important English case on the subject which identified certain powers whose “nature and subject matter are such as not to be amenable to the judicial process. The courts are not the place wherein to determine whether a treaty should be concluded or the armed forces disposed in a particular manner or Parliament dissolved on one date rather than another.”8 Nevertheless, the Court of Appeal had decided in Black that a prerogative power can be justiciable if it “affects the rights or legitimate expectations of an individual”. The Federal Court applied that test and found that the Prime Minister’s advice to dissolve Parliament did not affect such rights or expectations.
However, the Court did not stop there. It went on to consider, on the assumption that it had jurisdiction, whether the exercise of the prerogative of dissolution was “in accordance with the law” —in this case, section 56.1 of the Canada Elections Act. The Court quoted a passage from the Supreme Court of Canada’s decision in New Brunswick Broadcasting Co. v. Nova Scotia (Speaker of the House of Assembly):9
Our democratic government consists of several branches: the Crown, as represented by the Governor General and the provincial counterparts of that office; the legislative body; the executive; and the courts. It is fundamental to the working of government as a whole that all these parts play their proper role. It is equally fundamental that no one of them overstep its bounds, that each show proper deference for the legitimate sphere of activity of the other.
In that case, the Supreme Court had recognized the continuing vigour of the law of “parliamentary privilege”, and in doing so largely reaffirmed traditional limits on judicial intervention in the internal workings of the legislative branch. The Conacher case presented a potential clash between the judiciary and the executive branch. The Federal Court worried that, if the applicants’ interpretation was upheld, “…a court would be able to force the Prime Minister to dissolve Parliament, effectively dictating to the Governor General to exercise his or her discretion.” The Court was not prepared to go that far. Its conclusion was consistent with the conventional view that such a determination is essentially political, and that ultimately the electorate rather than the court is the proper forum to pass judgment on it.
Text of the Elections Law
The Federal Court proceeded, in any event, to consider the parties’ arguments over the actual language of the Canada Elections Act.10 The new section 56.1 of that Act reads as follows:
56.1 (1) Nothing in this section affects the powers of the Governor General, including the power to dissolve Parliament at the Governor General’s discretion.
(2) Subject to subsection (1), each general election must be held on the third Monday of October in the fourth calendar year following polling day for the last general election, with the first general election after this section comes into force being held on Monday, October 19, 2009.
A general election is formally launched by the Governor General issuing a proclamation under section 57 of the Act, fixing the polling day and requiring the distribution of writs of election.11 But an election of members to a newly constituted House of Commons cannot take place while the previous Parliamentis carrying on business as usual. Subsection (2) therefore presupposes that dissolutions of Parliament will occur on the same schedule as general elections.12 It is fundamental that enacted law such as the Canada Elections Act prevails over the common law, including the royal prerogative. Accordingly, it might have been be concluded, but for a provision like subsection (1), that the old discretionary prerogative of dissolution was also being supplanted by the new regime.
According to an old legal maxim, “Parliament does not speak in vain”, and so some meaning has to be attributed to subsection (1). A plausible reason for its inclusion is precisely to limit the application of the four-year rule in subsection (2).13 Counsel for the applicants did accept this up to a point, appreciating the need for discretion in the case of the government’s loss of confidence, but in no other circumstances. The respondents insisted that the Governor General’s power of dissolution remained unqualified.
It is a principle of statutory interpretation that laws are to be read in a manner consistent with the constitution. If a provision in an Act has more than one plausible interpretation, Parliament is presumed to have intended one that conforms with constitutional law, not one that would render the provision invalid. In this case, the rules governing constitutional amendment set forth in the Constitution Act, 198214 came into play. Parliament, acting alone, is empowered to amend the Constitution as it relates to “the executive government of Canada or the Senate and House of Commons”, but this is subject to certain exceptions. The Court considered that any diminution of the Governor General’s discretion would constitute an amendment of the “Constitution of Canada”15 and, moreover, one that affected “the office of the Governor General”. A constitutional amendment affecting that office (or that of the Sovereign) requires resolutions of concurrence by every provincial legislature. In the Court’s opinion, subsection 56.1(1) “explicitly leaves the Governor General’s discretion untouched” in order to avoid a conflict with the rules for constitutional amendment.
In support of the applicants’ restrictive interpretation of section 56.1, their counsel relied heavily on certain statements made by members of the government on the floor of the House and in committee. This “extrinsic evidence” was objected to by the respondents for the simple reason that it was at variance with other remarks made in the course of parliamentary debates. Citing examples, the Court found that this evidence taken as a whole was ambiguous and did not give it the authority to “read in” words of limitation that were not there. Such ambiguity in the parliamentary record is not particularly startling, as speeches are crafted differently and usually with less precision than the text of legislative enactments.
The Federal Court was also swayed by another practical difficulty that would have arisen if it had applied the applicants’ interpretation. Suppose a loss of confidence in the House of Commons were indeed a necessary condition for the calling of an early election. The courts would then be in the invidious position of determining when a loss of confidence occurs. There is no commonly agreed-upon definition of “non-confidence”16 against which a court could make an objective determination. In the words of the Court, “A government losing the confidence of the House of Commons is an event that does not have a strict definition and often requires the judgement of the Prime Minister.”
Fresh Conventions, Charter Rights
Unusual cases tend to inspire novel arguments. One of the more novel submissions in this case was that the enactment of section 56.1 had resulted in a change in the constitutional convention governing parliamentary dissolutions. It was argued that a new convention had been established, limiting elections to once every four years except in the case of a loss of confidence. The classical distinction between legal norms and conventional ones turns on the ability and willingness of the courts to determine and enforce them. However, in the 1981 Patriation Reference, the Supreme Court of Canada had agreed, in answer to a specific question, to determine the existence of a constitutional convention, at the same time acknowledging that there was no legal remedy available for a breach of it.17 Several commentators sounded the alarm over the potential implications of this precedent, including Justice B.L. Strayer in this passage from a book on judicial review:
Conventions govern many aspects of governmental activity and the potential which they create for judicial supervision of the political process seems immense. Will courts be called upon to determine under what circumstances a government can be considered to have the confidence of its Legislature, or when the Queen should act on behalf of her Canadian ministers? Consistently with the majority view in the Patriation Reference it would appear they can, although all of these issues involve what have normally been thought of as political criteria.18
Citing the Patriation Reference case as authority, the Federal Court considered whether the change in the election law had any impact on the accepted conventions surrounding dissolution. The extra-legal nature of conventions was noted and it was found that new ones were established by usage over time, not by legislative enactment at a point in time. As there had been no change in usage by the “relevant actors” (the Prime Minister and Governor General), the Court held the existing convention to be unaffected by the statutory amendment. Later, however, in summing up its findings on justiciability, the Court appeared to disavow any role in adjudging this question at all, saying that “the matter of convention… is political in nature and is outside the jurisdiction of the Court….”
The Charter of Rights, on the other hand, has opened myriad avenues of legal challenge to government action of all kinds, including on grounds that are indeed “political in nature”. The applicants here zeroed in on section 3, under the heading of “Democratic Rights”, which guarantees to citizens the right to vote in federal and provincial elections. Reliance was placed on precedents holding that the purpose of section 3 is to protect the right of citizens “to play a meaningful role in the electoral process” and to vote in an informed manner. The applicants contended that the Prime Minister’s use of his unlimited discretion to call the election differentiated between political parties in such a way to have an adverse impact on citizens’ ability to play their meaningful role. The Court rejected this claim which, in its view, would imply that every election since the Charter’s enactment in 1982 was in violation of section 3 as the government of the day, in each case, had exercised an unlimited discretion in the choice of a date.
What, then, are the legal (as opposed to political) effects of Bill C-16? Several commentators, writing before the Conacher case was decided, put forward interesting claims about its purpose and interpretation.19 The legislation was described as “purposely vague and problematically ambiguous”. It was said that the dissolution of Parliament in 2008, even if it fell within the literal terms of the law, ran counter to its spirit. And, according to one suggested interpretation, the general election envisaged for October 19, 2009 remained obligatory in spite of the intervening vote.
The early dissolution in September 2008, clearly permitted under subsection 56.1(1), started the four-year clock ticking a year early and so the default date was rendered superfluous. In the Federal Court, both sides agreed that an early election call was permitted by the law if occasioned by a non-confidence vote. It was during a minority Parliament, after all, that section 56.1 had been inserted into the Canada Elections Act. Members of that Parliament must have contemplated the possibility of an early loss of confidence, and therefore the chance of an election prior to October 19, 2009. Did they really intend that an election in 2008 must necessarily be followed by a second one a year later, even if the first election produced a majority government?
The essence of subsection (2) is to require that a general election be held not later than the month of October in the fourth year following the previous election. As previously noted, the Constitution of Canada fixes a parliamentary term at five years,20 but this is clearly an upper limit to ensure regular accountability to the electorate. Parliament does not offend that principle by legislating a shorter term within that maximum. Nor does the four-year maximum impair the Governor General’s “office” in the sense forbidden by the Constitution, since she retains the ability to dissolve a Parliament for as long as there is one in existence. On the other hand, sceptics might well argue, supported by some of the reasoning in Conacher, that the law lacks an effective means of enforcing a dissolution and election at the end of four years. On this view, the consequences of disregarding it would be political only.
Let us consider the analogous case of a Parliament that has carried on without a dissolution to the very end of the constitutional five-year limit. What would be the remedy in the unlikely event such a Parliament continued to meet and transact business, and the Speaker professed his continued authority to preside, and the Governor General failed to proclaim a dissolution on her own authority?21 The written Constitution does not expressly provide a formal mechanism for bringing proceedings to a close at that point. Would the courts feel constrained by the sacrosanct judicial rule against questioning “proceedings in Parliament”? It is submitted that they could find a remedy, not by intruding into the House’s internal affairs, but by refusing to give effect to legislation enacted by such a rogue Parliament. The grounds would be that the basis for Parliament’s existence, as laid down in positive law, had expired. If that is so, then can a similar analysis not be applied to section 56.1 of the Canada Elections Act, under which a “general election must be held” four years after the previous one?22 An answer by the courts to that question would determine whether the fixed date election law is a law in name only.
1. 2009 FC 920. A motion to advance the hearing to a day prior to polling day was denied in October 2008: 2008 FC 1119. The joint applicants were Democracy Watch, a public advocacy group, and its founder and coordinator Duff Conacher. The named respondents included the Prime Minister, the Governor General and the Governor in Council. The case is being taken to the Federal Court of Appeal.
2. S.C. 2007, c. 10, s. 1, amending S.C. 2000, c. 9.
3. Letters patent constituting the office of Governor General of Canada, effective October 1, 1947. They confer both general authority and specific authority (section VI) in connection with “summoning, proroguing or dissolving the Parliament of Canada”. See Maurice Ollivier (ed.), British North America Acts and Selected Statutes (1962). The Editor has noted that a government could simply amend the letters patent as an alternative means of qualifying the Governor General’s discretion.
4. The conventions of “responsible government”, by which the Crown when exercising its legal powers is bound by the advice of a ministry responsible to an elected House, were not firmly established in the U.K. until as late as the 1830s, but were effectively transplanted to most of British North America by 1848. See W.R. Lederman, “The British Parliamentary System and Canadian Federalism” (1971), reprinted in Lederman, Continuing Canadian Constitutional Dilemmas (1981).
5. R.S.C. 1985, c. F-7, s. 2, definition of “federal board, commission or other tribunal.” Although the definition includes powers conferred “by or under an order made pursuant to a prerogative of the Crown” this language does not extend to any of the prerogatives themselves. An example would be the order prescribing regulations for Canadian passports made under the prerogative power over foreign relations.
6. “Every House of Commons shall continue for Five Years from the Day of the Return of the Writs for choosing the House (subject to be sooner dissolved by the Governor General), and no longer.” The parenthetical qualification was made unnecessary by the drafters of its modern counterpart, s. 4(1) of the Constitution Act, 1982: “No House of Commons and no legislative assembly shall continue for longer than five years from the date fixed for the return of the writs of a general election of its members.”
7. (2001), 54 O.R. (3d) 215 (Laskin J.A.), on appeal from (2000), 47 O.R. (3d) 532 (Ont. S.C.).
8. Council of Civil Service Unions v. Minister for the Civil Service,  1 A.C. 374 (House of Lords).
9.  1 S.C.R. 319, 100 D.L.R. (4th) 212 (McLachlan J.)
10. The formulation of section 56.1 is not dissimilar from that found in several provincial regimes. Fixed-term election laws were adopted by British Columbia, Newfoundland, Ontario, New Brunswick and Prince Edward Island between 2001 and 2007.
11. See, generally, Patrick Boyer, MP, Election Law in Canada (1987), Chapter 3: The Calling of Elections.
12. It is a constitutional convention that the proclamation dissolving Parliament will be concurrent with proclamations respecting the issuance of writs of election and the meeting of a new Parliament: Robert Blackburn, “The Prerogative Power of Dissolution of Parliament: Law, Practice and Reform,”  Public Law 766. The proclamations issued on September 7, 2008 appear as an Extra number of the Canada Gazette (Part II), Vol. 142, No. 4.
13. Eugene Forsey, in typically colourful fashion, provides examples showing the incongruity of a “pure” U.S.-style fixed-term system and traditional parliamentary government: “Fixed Dates for Elections?” (1966), reprinted in Forsey, Freedom and Order (Carleton Library, 1974).
14. Constitution Act, 1982, ss. 41(a) and 44. Section 52 specifies what enactments are included in the “Constitution of Canada” for the purposes of that Act.
15. This position assumes that the “Constitution of Canada” includes prerogative powers (in addition to the enactments referred to in s. 52) or, alternatively, that the prerogative of dissolution has been codified in the 1867 Act (see note 6 above). The exact scope of the expression has been the subject of lively debate: see Warren J. Newman, “Defining the ‘Constitution of Canada’ since 1982”, Supreme Court Law Review (2d series), vol. 22, p. 423 (2003) at pp. 476-484.
16. On this point generally, see Eugene Forsey, “The Problem of ‘Minority’ Government in Canada” (1964), reprinted in Forsey, op. cit., note 13 above.
17. Re Resolution to amend the Constitution,  1 S.C.R. 753; 125 D.L.R. (3d) 1. The questions referred to the Court concerned the extent of provincial consent required, by law or convention, for constitutional amendments sought by Canada from the U.K. Parliament up to 1982.
18. The Canadian Constitution and the Courts (3rd ed., 1988), at p 232.
19. See, for instance, Guy Tremblay, “The 2008 Election and the Law on Fixed Election Dates,” Canadian Parliamentary Review, Vol. 31, No. 4, Winter 2008; Adam Dodek, “Fixing Our Fixed Election Date Legislation,”ibid., Vol. 32, No 1, Spring 2009.
20. Note 7 above. However, under s. 4(2) of the Constitution Act, 1982, a law enacted with the support of two-thirds-plus of the members of the House of Commons may extend a Parliament beyond five years in time of “real or apprehended war, invasion or insurrection.”
21. Constitutional scholar Peter Hogg, in response to questions in the Standing Senate Committee on Legal and Constitutional Affairs on Feb. 8, 2007, was of the view that the Governor General would not act without advice to limit the duration of a Parliament pursuant to s. 56.1. See the committee’s Proceedings, Issue 21, 1st Session, 39th Parliament.
22. On the one hand, the Constitution unlike the elections law is the “supreme law of Canada” (relative to other laws when they are in conflict). On the other hand, the five-year limit in s. 4 of the Constitution Act, 1982 could arguably be made inapplicable to the House of Commons by an ordinary Act of Parliament (pursuant to the s. 44 amendment procedure) and in that sense it enjoys no higher a status than s. 56.1 of the Canada Elections Act.