'Harper's new rules' for government formation: fact or fiction?

• Author: Knopff, Rainer
• Position: Stephen Harper - Essay

Prime Minister Stephen Harper is said to have taken a new and constitutionally suspect approach to government formation in 2008, insisting that only new elections could change parliamentary governments. "Harper's New Rules," (1) generated an outpouring of criticism from constitutional scholars.

We challenge the critique of "Harper's New Rules" primarily as it appears in two of its leading exemplars: the work of Peter Russell, a constitutional scholar of high repute whose writings always merit careful consideration, and of the late Peter Aucoin, Mark D. Jarvis and Lori Turnbull (cited hereafter as Aucoin) co-authors of Democratizing the Constitution a fine, prize-winning book on responsible government. (2) Even the best authors and books are open to question and debate, as we think both Russell and Aucoin are with respect to "Harper's New Rules."

Background and Context

When facing possible defeat in the Commons just six weeks after being re-elected with a strengthened minority government in 2008, Stephen Harper asserted that a proposed coalition of the Liberals and NDP (with the promised stable support of the Bloc Quebecois) could not legitimately be appointed as an alternative government by the governor general, even at this very early stage of the Parliament's life. While "the opposition has every right to defeat the government," Mr. Harper maintained, "Liberal leader Stephane Dion does not have the right to take power without an election. Canada's government should be decided by Canadians, not backroom deals. It should be your choice--not theirs." (3) Days later, in a heated parliamentary debate with Mr. Dion, he claimed "the highest principle of Canadian democracy is that if one wants to be prime minister one gets one's mandate from the Canadian people and not from Quebec separatists." (4) The Prime Minister's statements have been widely understood as meaning that defeat of a plurality minority government must always trigger new elections, because only elections can legitimate a new government. For Harper, says Peter Russell "the only way to get rid of a government that does not have the confidence of the House of Commons is to elect another House of Commons." (5) Aucoin agrees, "If Harper's view were to be accepted," he maintains, "the only option would be ... dissolution and an election to choose a new House after every loss on a confidence vote." (6) Of course, requiring an election after every loss of confidence or as the only way of replacing a defeated government raises the spectre of a "diet of successive elections in a short period of time." Especially in circumstances of the kind of "fragmented electorate" that generates minority governments, says Russell, Harper's approach could mean being "bombarded by an unending series of elections until one party secures a majority." (7)

Such an "elections-only" theory of governmental change, its critics rightly insist, is inconsistent with well-established practices of our system of parliamentary government, namely, that the governor general may appoint an alternative government following a vote of non-confidence, at least early in the life of a parliament. "Until recently," writes Aucoin, "most experts would probably have agreed that the governor general could properly refuse the prime minister's advice for a dissolution following the government's defeat on a confidence vote in the House of Commons if the loss of confidence came shortly after an election." (8) But this established consensus is now eroding in the face of contrary opinions, led by Prime Minister Harper. In other words, a previously dominant viewpoint is under challenge from the novel and disturbing theory known as "Harper's New Rules."

Because Harper's own statements concerning these "new rules" are brief and made in the heat of political battle, Russell indicates that it would be better to call them the "Harper/Flanagan rules, because political scientist Tom Flanagan, a long-time adviser of Harper," provided a more "extensive elaboration" of them than Harper himself. Aucoin similarly associates Flanagan with the view that "elections should be the only way to change a government from one party to another" and attribute to him the "scholarly rhetoric" in support of Harper's constitutionally suspect elections-only theory of governmental change. Other authors have also seen Flanagan as best reflecting and explaining Harper's views on this issue. (9) While Flanagan is universally considered the chief theoretician of "Harper's New Rules," Aucoin considers Michael Bliss and Andrew Potter as providing additional support for this position.

Misreading Flanagan, Bliss, and Potter

Those who attribute to Flanagan the view that loss of confidence must always trigger new elections rely exclusively on a gingle Globe and Mail op-ed, published on January 9, 2009. Aucoin clearly sees this article as justifying the elections-only position on governmental change (10) and reproduces much of the article on pages 175-76 of Democratizing the Constitution. A key notion in Flanagan's piece is that "the most important decision in modern politics is choosing the executive of the national government, and democracy in the 21st century means the voters must have a meaningful voice in that decision." (11) This view, says Aucoin, "empowers parties, rather than individual MPs or Parliament as a whole, by seeing government as the entitlement of the party that has won the most seats," and it "follows logically.., that if the party loses confidence, then the people must elect a new 'party'," (12) or, as Russell puts it, "that the prime minister cannot be changed without another election being called." (13)

But that position--which the critics clearly ascribe to Flanagan--is difficult to square with another Globe and Mail piece that Flanagan had published just one month earlier, a contribution that none of the critics we are considering (so far as we can tell) ever acknowledge. (14) The December 2008 article, entitled "This coalition changes everything," provides the following answer to the question whether the governor general should grant the likely request for a new election upon defeat by the coalition:

Normally, the question would be easy to answer. Since the last election was so recent, a defeated prime minister should not expect a new election, and the opposition should get the chance to govern if it can offer a plausible plan for stability, which the opposition has done with its proposal for a Liberal-NDP cabinet supported by the Bloc. (15) This is very far from saying that if the governing "party loses confidence then the people must elect a new 'party'." Indeed, it explicitly acknowledges that "normally" this should not occur early in a parliament's life, that "normally" the governor general should in such circumstances refuse a dissolution request and appoint an alternative government. We have emphasized the use of the word "should" in Flanagan's piece in order to underline how thoroughly it fits into the older consensus represented in the upper right cell of the table on the following page, perhaps as far to the right as Eugene Forsey, who the table describes as thinking not only that refusing dissolution is constitutionally permissible but also that it should occur under specified conditions.

Flanagan's December 2008 piece explicitly invokes Forsey in support of his view that dissolution should not "normally" be granted early in a parliament's life. However, he then goes on to rely on the same Forsey to conclude that the 2008-09 situation is not "normal."

But this is not a normal situation. Constitutional expert Eugene Forsey famously supported Lord Byng's refusal of Mackenzie King's request for an election in 1926, but even Mr. Forsey had to admit that an election would have been necessary if "some great new issue of public policy had arisen, or there had been a major change in the political situation." (17) For Flanagan, "the emergence of the opposition coalition has satisfied both of Forsey's conditions for going back .to the voters." (18) It did so for two reasons: first, because the coalition relied on the promised stable, ongoing support of a separatist party; second, because key participants in the coalition had explicitly rejected the very idea of a coalition during the just completed election campaign. (19) Flanagan's 2008 op-ed clearly did not take the general position in favour of new elections every time a minority government was defeated on a confidence vote that critics ascribe to the piece he wrote just one month later. Moreover, the clear implication of his 2008 piece is that if a new election this time fought with the possibility of coalition obviously in mind--returned another Conservative plurality, the opposition parties should expect the governor general to appoint the coalition if it defeated the Conservative government on a confidence vote soon after the election.

It is possible, of course, that Flanagan changed his mind over the course of that month, and that he had moved to the "elections only" position by January 2009. But there is evidence that he had not fundamentally changed his mind. Part of that evidence is found in the 2009 article itself, in passages that Flanagan's critics never quote (just as they systematically ignore his 2008 piece). Even Aucoin, who reproduces almost everything else in the January 2009 article, leaves out the part in which Flanagan elaborates his statement that "gross violations of democratic principles would be involved in handing government over to the coalition without getting approval from voters":

Together, the Liberals and the NDP won just 114 seats, 29 fewer than the Conservatives. They can be kept in power only with the support of the Bloc, whose raison d'etre is the dismemberment of Canada. The Liberals and N...