As the parties grew stronger, few candidates could hope any longer to be elected without the support of these loyal party workers and the party machine. The price of that support was loyalty. Soon, says historian Noel, Mowat “had no need to bargain in the lobbies of the legislature with capricious independents and local patrons who would never lend more than their conditional or nominal support to any leader.”36
Yet even with their machines primed and their caucuses tamed, leaders like Mowat and Macdonald did not enjoy unchallenged authority. Party leadership could still be removed at a moment’s notice. In 1880, Alexander Mackenzie, former prime minister and leader of the Liberal opposition in Ottawa, was abruptly replaced when his backbenchers decided they preferred Edward Blake to lead them. In 1896, the Conservative cabinet forced the resignation of Prime Minister Mackenzie Bowell – “a weak, vain, decent old mediocrity,” in historian Peter Waite’s phrase, which seems to be the kindest thing anyone has said about him – in the hope that Charles Tupper could save them from Wilfrid Laurier’s Liberals. But the era of independent members who could topple a government or unseat a leader was waning. Far-sighted leaders were using the party machines to drain the pool of loose fish.
Disciplined party organizations flexed their muscles in all democratic societies late in the nineteenth century. In Britain, Liberal leader William Gladstone invented “the platform” and began to run national campaigns focused on the leader. Soon the British Labour Party demolished the cosy, gentlemanly politics of “the ten thousand” through national campaigns aimed at the masses of working-class voters who had finally been enfranchised. (Walter Bagehot, who died in 1877, was spared from seeing it.) In Canada, Macdonald’s “National Policy” platform also directed attention away from local candidates to the party and its leader. It was after Macdonald’s death, however, that Canada became unique among parliamentary democracies in the steps it took to reduce the ability of backbenchers to influence their leaders.
In 1919, the Liberal Party, split by the conscription crisis of 1917 and left leaderless by the death of Wilfrid Laurier, turned a policy conference it had planned into a leadership convention. William Lyon Mackenzie King wrapped himself in the mantle of Laurier, won the race, and became the first Canadian party leader chosen by a party convention rather than by the parliamentary caucus. The party hailed the convention that chose Mackenzie King as a great advance over the older method, and every other party followed its example. Since 1919, federal and provincial parties in Canada have chosen their leaders in mass party gatherings.*
Canadians have celebrated leadership conventions as a way to let the people participate directly in government, and the selection process has expanded steadily. Instead of being gatherings of a few hundred, dominated by the party brass, conventions came to involve thousands of delegates choosing among candidates who spent millions of dollars on elaborate media campaigns. In the 1980s, the parties began to abandon conventions themselves in favour of even more broadly based processes, in which a hundred thousand or more party members could cast leadership ballots.
Canadian politicians and analysts have always identified mass membership participation in the choice of party leaders as a triumph of democracy. The cost, barely noticed, was the abrupt loss of what little influence elected members of Parliament still had over their leaders. Mackenzie King understood his new authority perfectly. He never sent his cabinet colleagues to the penitentiary, but he sometimes kept their signed resignation letters on file, and he used one when his defence minister tried to stand up to him in the conscription crisis of 1944. King emphasized to his caucus as often as necessary that, since it could not challenge his leadership, it could not challenge his policies. He represented the party and was answerable to it, not to them, and he controlled his MPS ruthlessly.
Being “answerable” to the party was no great burden to King; the Liberal Party did not meet again in convention until he chose to retire in 1948. King had made “Parliament will decide” his maxim, and he trotted it out whenever he wished to avoid a decision. He knew that, so long as his party won the general elections, Parliament would decide what he told it to. Since the other parties still acted as if Parliament was the essential forum, King’s Liberal Party enjoyed remarkable success for a generation. Macdonald would have admired, and envied.
King actually had substantial parliamentary skills and might have prospered as a party leader answerable to caucus. John Diefenbaker was the first Canadian to become a party leader and a prime minister on the strength of a “grass-roots” leadership campaign. Sweeping to the leadership of his party and then the country on the strength of his personal appeal to voters. Diefenbaker soon demonstrated his inability to work within the machinery of parliamentary government. Since his cabinet and caucus had no authority to challenge him, Diefenbaker’s party was thrown into years of turmoil when he refused to relinquish the party leadership, even in defeat.
Still, the principle that Canadian party leaders should be immune to parliamentary control survived. In 1988, when the opposition Liberal caucus members in Ottawa had the same kind of dissatisfaction with their leader as their ancestors had felt with Alexander Mackenzie in 1880, a majority of them signed a declaration that John Turner must resign. Unlike Mackenzie, however, Turner could defy the members of Parliament. Claiming “democratic legitimacy,” because he had been chosen by a leadership convention, Turner obliged the dissident majority to recant. Turner led his party into the 1988 election and resigned only after the party’s defeat.
The independence of Canadian party leaders has been most spectacularly underlined in provincial, rather than national, politics. William Vander Zalm, elected premier of British Columbia soon after winning a leadership convention in 1986, ran an erratic one-man government that alienated many of his cabinet and caucus members, few of whom had wanted him as leader. As he became mired in scandals in 1991 and fell precipitously in the polls, virtually all wanted rid of him. All accepted, however, that Vander Zalm was not answerable to them. They dutifully supported him in office until a conflict-of-interest commissioner (an unelected civil servant personally appointed by the premier) told Vander Zalm he must go. Almost all his party’s elected members lost their seats in the election that followed.
By the late twentieth century, it had become understood in Canadian politics that backbenchers were responsible to their leaders, rather than vice versa, even when blind allegiance ensured their imminent defeat. This faith in the leader’s unlimited authority was as strong in new “grass-roots” parties as in established ones, and it was accepted in regional blocs as well as national parties. Even the occasional proposals to “strengthen” Parliament, usually by encouraging party leaders to “allow” more “free” votes on matters of limited importance to them, confirmed the absolute power of leaders over their followers’ votes. By the time Pierre Trudeau joked that members of Parliament were nobodies fifty yards from the House of Commons, it was unclear why he would invoke the fifty-yard rule.* In the name of participatory democracy, the elected representatives of the Canadian people had become the only individuals in the country without any political opinions of their own.
When general elections in Canada cost $100 million, almost all of it funnelled through the political parties (heavily subsidized by taxpayers), virtually no one could be elected except on the parties’ terms, and most proposals to change the system, through proportional representation and “recall” initiatives, for instance, were aimed at weakening individual members even more. In other countries with parliamentary governments, the power of political parties – and their leaders – had also expanded greatly. But from Australia to Japan, and almost everywhere else that the parliamentary system endured, elected representatives still retained a veto over party leadership and party policy.
Bob Hawke launched his career as the most successful prime minister of recent Australian history in 1983 when the Australian Labour Party caucus, worried about its electoral chances, dumped its leader and chos
e Hawke in a hastily convened meeting on the day the Conservative government called a federal election. Hawke went on to win the election and to dominate Australian politics for a decade. In 1992, however, the caucus that had chosen him decided on the eve of another election that Hawke, the sitting prime minister, had become a liability. Leading the challenge to Hawke was Paul Keating, a rival who had left Hawke’s cabinet to contest the leadership from the backbenches. Even though Australia was much less regionally diverse than Canada, Hawke and Keating fought their battle by seeking support from the powerful regional blocs within the Labour Party caucus. Keating won. He replaced Hawke as party leader and prime minister, and defied the polls by leading Labour to an upset majority victory.
In New Zealand, it was policy struggles that fuelled leadership battles. In the 1980s, David Lange’s opposition to nuclear testing in the South Pacific made him perhaps the only world-renowned prime minister New Zealand ever had. But as New Zealand’s deficit worsened, Lange and his finance minister became embroiled in economic policy disputes. Unable to hold his caucus’s support on a matter of fundamental party policy, Lange was bounced from the party leadership and the prime ministership.
The most famous beneficiary and victim of caucus control over parliamentary leadership was British Prime Minister Margaret Thatcher. As a woman and as a politician with a combative personality and strong policy views, Thatcher was the kind of candidate who could never have won a leadership convention. Caucus colleagues identified her leadership potential, and their support enabled her to depose an incumbent leader and go on to lead the British Conservative Party to three consecutive election victories. Then, while Thatcher was still at the peak of her influence as a world leader, her caucus dumped her as swiftly and ruthlessly as it had chosen her. By then, the British Labour Party had moved to the Canadian system of letting the party at large select party leaders. Its leader, Neil Kinnock, was described by veteran British political commentator Peter Jenkins as the kind of leader “who could not conceivably have come to the fore” except by leadership convention.38 Against him, Thatcher’s replacement, John Major, won an upset victory.
The crisis that had provoked Britain’s Conservative Party caucus to depose a sitting prime minister was a deeply unpopular new tax, the “poll tax.” Prime Minister Thatcher had been determined to implement it in defiance of the public mood, even though most of her backbenchers feared it would make their re-election impossible. Deposing Thatcher enabled the backbenchers to kill the poll tax, and enough of them survived the ensuing election to keep the Conservatives in power. Their aim was saving their seats, that is, but the result was that a prime minister with a majority government had been prevented from implementing an unpopular policy.
Almost simultaneously, the Canadian prime minister, Brian Mulroney, introduced his own unpopular taxation measure, the Goods and Services Tax. It, too, helped provoke a disastrous slide in party popularity, but Mulroney’s backbenchers understood they had no right to interfere in either policy issues or party leadership. They dutifully voted to impose the new tax. Mulroney retired at a time of his own choosing. His hapless backbenchers, obliged to defend the new tax under a new leader chosen by a leadership convention, suffered annihilation in the 1993 election.
John A. Macdonald was joking – and understood to be joking – when he told Robert Dickey he wanted backbenchers who would support him even when they thought he was wrong. Macdonald would have enjoyed such support, and he worked diligently to bring about a rough approximation of it. But he never lived in an era of absolute party leaders. He was a brilliant parliamentary leader because his prime ministership was always subject, not only to ratification at general elections, but also to constant, critical, informed parliamentary review of a kind his late-twentieth-century successors never had to fear. That is, Macdonald as prime minister was responsible to a parliamentary majority – rather than it being responsible to him.
This reflection on parliamentary leadership in a book about the making of confederation in the 1860s has seemed worthwhile because the changed nature of leadership has constitutional significance. Despite the late-twentieth-century consensus that confederation was imposed by an undemocratic and anti-democratic élite, confederation in the 1860s was not a prime ministerial measure. It was not a party measure. It was not a cabinet measure. It was, above all, a legislative measure.
In order to woo sceptical, independent-minded legislators, confederation’s advocates in 1864 had been obliged to recruit bipartisan delegations. These delegations engaged in heated debates on the way to a consensus they thought might win the support of their legislative colleagues, even though several of the delegates, as well as most of the excluded political factions, continued to dissent. The confederation proposals that emerged at the conferences were vigorously tested in all the legislatures, in debates that were listened to because every legislator was potentially a recruit to either side. “Loose fish” did triumph over party leaders, at least temporarily, in all four Atlantic provinces. They did not in either Canada West or Canada East, but they could have if confederation had not satisfied them, and leaders like Hector Langevin feared that they might.
The ratification of confederation in British North America between 1864 and 1867 was a deeply imperfect process. An electoral system that excluded women for their gender and men for their lack of property was far too narrowly based. Substantial, politically active minorities, such as the Acadians of New Brunswick and the Catholics of several provinces, were shockingly under-represented in the process. Aboriginal nations and racial minorities such as black and Chinese Canadians were almost completely disenfranchised. All parties applied crude and gamy techniques of persuasion to legislators and voters alike, and the system provided few opportunities for redress. No one spoke of a charter of rights.
Nevertheless, amidst the heated battles, the furious allegations, and the deep-rooted opposition that confederation provoked during the 1860s, the Canadian constitution that came into effect on July 1, 1867, had been judged – and finally endorsed – by legislatures of elected representatives who maintained substantial independence from, and ultimately a veto upon, the governments that were responsible to them. Doubts, resentments, and even deep opposition remained. But the process by which the new constitution had been negotiated and ratified had conferred on it a legitimacy that was celebrated by its supporters, acknowledged grudgingly by opponents, and – above all – generally accepted by the Canadian population that would begin to govern itself under it.
Superficially, constitution-makers had things much easier in the 1980s and 1990s. To produce the British North America Act for five small colonies with barely three-and-a-half-million people in the 1860s, thirty-six mutually hostile and politically diverse delegates had to engage in weeks of intense negotiation, and their draft agreement was subjected to two years of vigorous, open-ended, and unpredictable legislative debate. By comparison, the Meech Lake accord of 1987, drafted by lawyers and bureaucrats, required barely a day of discussion by the first ministers – the famous “eleven white men in a room.” Slated for ratification by docile legislatures almost entirely subservient to executive control, it was derailed almost by accident. The patriation round of 1981 had been similarly exclusive, and Quebec in particular never accepted the legitimacy of the process, which produced what came to be known as “the Trudeau constitution.” The process leading to the Charlottetown accord of 1992 included more public comment, but its final terms were once more a matter of executive decree. The 1992 referendum, in which voters were invited either to ratify the accord or be held liable for the destruction of the country, reflected the widespread belief that “democracy” and “direct democracy” were the same thing – and Parliament largely irrelevant. Though it ignited passionate debate among Canadians, the referendum was a very limited substitute for genuinely representative scrutiny of the terms of the deal.
In the late twentieth century, Canada continued to govern itself through parliamentary forms.
But the long quest for more “direct” participation in government had – paradoxically – conferred enormous power upon first ministers. It was not that the politicians sought to be dictators. They submitted to leadership conventions, fought election campaigns, occasionally held referendums, and followed the polls devoutly. But they and their advisors made constitutions in a vacuum, because the only ongoing, effective control upon them – control by legislatures – had largely ceased to operate. Parliament no longer provided a forum in which elected representatives could test and potentially reject executive initiatives. Except for the election-night tallying of which party leader controlled the most seats and would become prime minister, a seat in the Commons had become largely ceremonial. Dignity was not a word much associated with late-twentieth-century Canadian parliaments and parliamentarians, but Walter Bagehot would have said that they had indeed shifted from the efficient to the dignified side of the constitution.
Would John A. Macdonald, that connoisseur of power, have envied the vast freedom conferred upon his successors as prime ministers and premiers? Perhaps not. Macdonald, like all the confederation-makers, had undergone his apprenticeship to politics in the moment when “responsible government” was shiny, fresh, and new. He wielded executive power ruthlessly when he could, and he always sought to maximize it. But in his day, elected legislators had only recently wrested power from unelected, arbitrary executives. They guarded their new responsibility jealously, and they used it both to make or unmake governments and to veto or ratify policies. No policy could achieve legitimacy without their consent. In nineteenth-century Canada, the great convenience of ruling by executive decree was understood to be cancelled out by its lack of legitimacy – as Oliver Mowat demonstrated to Macdonald so convincingly in the fight over the power of disallowance. By the late twentieth century, that understanding had largely been lost.
1867 Page 25