The Canadian party system baffles outsiders and Canadians alike. Outsiders either cherry-pick details to suit a theory or just throw up their hands. Canadians wear the idiosyncrasy as a badge of honour; when looking for explanations they rarely look beyond the case. In The Canadian Party System: An Analytic History (UBC Press), I try to do the case justice from both a Canadian and a comparative perspective.

The system is anomalous in several ways. It defies the most powerful generalization in empirical political theory, Duverger’s Law. The law states that in a strongly majoritarian institutional context such as ours — most importantly, with votes counted within single-member districts — voters will concentrate on two and only two parties. Instead, Canada has had a multiparty system since at least the 1930s.

The system has spasms of massive volatility. Even though the dominant parties are among the oldest in the world, each has been to death’s door and back. In some provinces, the same citizens support fundamentally different party systems in provincial elections than in national ones. Although a party of labour exists (the NDP), the system’s class basis is weak. For most of the 20th century Canada’s political gap between Catholics and Protestants was as wide as any gap anywhere.

Finally, unlike most systems — and unlike all other systems with single-member districts — the Canadian system is dominated by a party of the centre. The Liberal Party of Canada manages to control the centre both on the classic left-right dimension and on Canada’s own “national” question. The anomaly of dominance from the centre is the key to understanding most of the other anomalies in the Canadian party system and yet begs explanation itself.

The place to start, then, is with the Liberal Party. Its dominance of Canada is a function of its dominance of Quebec. The Quebec electorate has been remarkably coordinated, both in concentration on one party at a time and in mobility between parties. These traits are then amplified by the electoral system. The rest of the country has rarely matched Quebec in electoral coherence, however. Quebec alone would put the Liberals halfway to a parliamentary majority.

What Quebec gives, it also takes away: control of Quebec was not just a sufficient condition for power, it was also a necessary condition. If the Liberal Party failed to hold a large Quebec bloc, it lost power. Sometimes, Quebec would go all the way to the other side and give the Conservatives a head start, as happened in 1984. If Quebec’s history of concentration accounts for Liberal dominance, its mobility accounts for much of the party system’s history of flux.

If a party survives as centrist on one dimension of debate, it must be off centre on some other dimension. So theory says. How can a party survive at the centre on both axes, both the left-right axis and that of the national question? The answer is that on the national question the Liberals are centrist only in a residual sense. The party occupies one pole inside Quebec and the opposite pole elsewhere. Quebecers supported the Liberals as the only acceptable federalist alternative. This left plenty of room on the nationalist side of the province’s electorate. Most of the time no party filled that nationalist space, but when one did, the Liberals stood revealed as only modestly pro-Quebec — and as vulnerable. Outside Quebec, in contrast, the Liberals were usually the most pro-Quebec party, always with another party with an anti-Quebec bent. It was by taking Quebec and the rest of Canada together that the Liberals held the existential centre.

When culturally charged questions heated up, as happened with Canada in the British Empire early in the 20th century or with Quebec in Canada later in the century, the Liberals were vulnerable on both flanks. All elections since 1993 are cases in point, but so were several before then. Most notable were 1911, 1958, 1984 and 1988, when the Conservatives, the anti-Quebec party of the 20th century, seized the pro-Quebec position inside that province. Such ends-against-the middle groupings are hard to sustain in multiparty coalitions, but they are next to impossible to sustain — much less to assemble — within the bosom of a single party. When this happens, the pattern is boom and bust, another component of Canada’s electoral volatility.

For most of the 20th century, the centrality of Quebec made cultural identity important in the rest of the country, with Roman Catholics of every ethnic background the province’s key allies. This was not because Catholics were always sympathetic to Quebec or francophone interests. The commonality, rather, was indifference or resistance to a British conception of Canada. For the first half of the 20th century the core of identity conflict was over how much blood and treasure Canada owed the empire and, relatedly, how aligned Canada’s external policy should be with Britain’s. The division emerged in its 20th-century form in the 1900 election, which was fought over the South African War. The conscription crisis of 1917 blew the gap wide open, and the division persisted into the 1960s.

Anti-conscription parade at Victoria Square, Montreal. May 24, 2017. Library and Archives Canada / C-006859

By the 1960s, external policy had been displaced by conflict over the symbolism of the state, with the flag debate as the high point. Once that was settled, the gap between Catholics and others began to shrink, and it has been nonexistent for several years now. As the overarching cultural difference dissolved, identity politics shifted to group-specific appeals, including the question of Quebec in Canada, with complex and shifting coalitions.

The inner workings of government
Keep track of who’s doing what to get federal policy made. In The Functionary.
The Functionary
Our newsletter about the public service. Nominated for a Digital Publishing Award.

Many Canadians do not want to be forced into any culturally defined choices — French versus English, Catholics versus Protestants — and resistance to a cultural agenda is greatest in the West. As westerners try repeatedly to shift the agenda, the result is fragmentation of the party system. New parties enter the system but the old ones refuse to leave. This fragmentation has taken two forms, each with its own geography and with different implications for the rest of the system.

One kind of new party might be called “insurgent.” These are parties that typically appear suddenly, have geographically concentrated support, enjoy short shelf lives and disappear as suddenly as they appear. The Progressives were like this in the 1920s, and so was the most dramatic recent arrival, Reform (later the Canadian Alliance). Insurgents increased the territorial component of the vote, creating sharp contrasts between East and West and, later, between Quebec and the rest. These contrasts were accentuated by the electoral system.

In the early days insurgents were nonideological, or at least featured programs that were not on a left-right axis; often they signalled economic distress. The Progressives were like this, both in that their program was focused on farmers and in that they surged with the depression that followed the First World War and declined with the recovery in the early 1920s. Later insurgents, in contrast, anchored the right side of the ideological spectrum. Insurgents became the dance partners for Conservative boom and bust, alternating between stealing support from that party and giving the support back. The Reform Party aside, insurgents tend to win bigger and last longer in provincial elections than in federal ones, and this accounts for much of the divergence in some provinces between the federal and provincial party systems. These are stories mainly about BC, Alberta and Quebec.

The shift of insurgents to the right was in reaction to the appearance of a qualitatively different new party on the left, first the Co-operative Commonwealth Federation (CCF), then the New Democratic Party (NDP). Although the CCF was founded by a mix of urban ideologues, farm organizations and high-minded reformers, its professionals pushed from the start for it to become a party of organized labour. Realizing this ambition required shifts elsewhere on the landscape, most importantly the merger of craft and industrial unions into the Canadian Labour Congress. The NDP emerged on the scene, and just as insurgents on the right increasingly challenged the Conservatives, the new party on the left focused on the Liberals. But where the insurgent story on the right is episodic, that of the NDP has been more of a slow trend, rarely boom or bust. And the NDP, although originally strongest in the West, gradually spread eastward. As it does so, competition within each riding has been increasingly multiparty.

Tommy Douglas and the CCF parliamentary group, 1942. C.C.F. / Library and Archives Canada / C-000314

Multiparty competition within ridings is where Canadian political reality proves comparative politics theory wrong. Whatever happens in the country as a whole, multiparty contestation involving a party of labour is not supposed to take place locally. In other countries, the nonsocialist rivals become one party, or nearly so. Why, then, has the NDP not made more inroads? Why, alternatively, does it persist in futility? The answer to both questions is the continuing strength of the Liberal Party. This strength is not just the mirror image of NDP incompetence, real or imagined. Rather, it is — or was — based on the specifically Canadian story of Quebec, the nation within a nation and its historic record of survivance through one-sided support of the Liberals. At the same time, the fact that the usual Canada-wide winner was the Liberal Party allowed the NDP to hang on: if the left was not allowed to govern, neither usually was the right. The exceptions — recent swings against the NDP — are the exceptions that prove the rule: they occur when it is time to show Conservative governments the door.

One of the ways I illustrate this logic is by looking at an electoral arena where many Canadians behave exactly as comparative theory says they should: in the provincial arena in the West. Here, antisocialist consolidation is a way of life, as Alberta has just shown with the recent merger of the Progressive Conservative and Wildrose parties into the United Conservative Party. Also, as in other countries, the identity of the party on the right is a historical accident, reflecting earlier political developments in each province’s history. In Manitoba and Alberta, the party on the right carries the label Conservative. In BC, it is the Liberal Party. In Saskatchewan, it is the Saskatchewan Party. The pattern in these provinces — the NDP as a constant on the left and a province-specific nonsocialist rival on the right — uncannily replicates the party history of countries such as Britain, Australia and New Zealand. This pattern also helps explain the divergence between federal and provincial party systems.

My agenda in The Canadian Party System was not just to try to make sense of Canadian complexity but also to shed light on controversies and claims about the rest of the world, especially the English-speaking world of single-member-district systems. In Canada, the old coexists with the new; different parties (quite sensibly) obey different logics; competitive contexts can vary as much within provinces as between them (again for very good reasons). These patterns generate ongoing motion and make the Canadian electorate a living laboratory, a replication in microcosm of the rest of the world.

Photo: Federal party leaders (left to right) Jean Chretien, Jean Charest, Preston Manning, Alexa McDonough and Gilles Duceppe chat briefly before the federal debate in Ottawa in 1997. The Canadian Press, Tom Hanson.


Do you have something to say about the article you just read? Be part of the Policy Options discussion, and send in your own submission. Here is a link on how to do it. | Souhaitez-vous réagir à cet article ? Joignez-vous aux débats d’Options politiques et soumettez-nous votre texte en suivant ces directives.

Richard Johnston

Richard Johnston is the Canada Research Chair in public opinion, elections and representation at the University of British Columbia.

You are welcome to republish this Policy Options article online or in print periodicals, under a Creative Commons/No Derivatives licence.

Creative Commons License

More like this