When he called the election this past Sunday, Stephen Harper entered his fifth campaign as leader of the Conservative party. Harper has had a great deal of success, over five elections he has taken the right in Canada from being divided into two parties and stuck in opposition to minority and then majority governments. The rise of the Conservatives has been paralleled by a collapse in Liberal support. The party that dominated the 1990s (and much of the 20th century) has experienced declining vote shares through the 2000s to the point where in 2011 Canada’s former “naturally governing party” fell to third place. Understanding past elections is important to understanding current ones. A dramatic Liberal decline that occurred steadily over the course of a decade cannot be explained by looking at one or two elections. One has to examine why the Liberals were so successful in the decades prior to the 2000s and what is different now.
Through most of the 20th century the base of Liberal support was in Quebec. In the first half of the 20th century Liberals governments would win by linking support in Quebec with support in Western Canada and to a lesser extent Ontario. In the later half of the century, after the rise of Diefenbaker and the building of a strong Conservative base in Western Canada, Liberal victories came from an alliance of Ontario with Quebec. This changed in 1984. In that year’s election Mulroney’s Conservatives took 58 seats in Quebec compared to the Liberals 17. The Conservatives built on that total in 1988, increasing their share of Quebec seats to 63 (as compared to 12 for the Liberals). Indeed, the Liberals have not managed to win a majority of seats in Quebec since the 1980 election. When the Conservatives collapsed in 1993 they were replaced by the Bloc Quebecois, and when the Bloc collapsed in 2011, they were replaced by the NDP. The loss of the Liberals’ Quebec base should have meant disaster for the Liberal’s governing prospects but in the 1990s they strung together three straight majority governments for the first time since the 1950s. Vote splitting on the right of the political spectrum between the Reform/Canadian Alliance and the Progressive Conservatives is often considered a partial explanation for the Liberal success in the 1990s. An examination of what would have happened had the right not split, though, suggests that Liberals would have thrived under all but the Conservative’s most optimistic scenarios.
In order to test what might have happened had the Liberal run against a united right in 1993, 1997, and 2000 I calculated results for an imaginary united Conservative party on a riding by riding basis under several different scenarios. The first is a naive scenario in which all Reform/Canadian Alliance and Progressive Conservative voters support the Conservatives. This required simply adding the Reform/CA vote totals to the PC vote totals in each district to generate a vote share for the united Conservatives or CPC. This scenario is, however, unlikely to have been the reality had a single Conservative ran in the 1990s. There were undoubtedly Progressive Conservatives who felt closer to the Liberal party than to the Reform and Canadian Alliance. Indeed, when the parties merged 4 of the 12 PC MPs (including Joe Clark and Scott Brison) refused to join the new party. Brison would cross the floor to become a Liberal cabinet minister under Paul Martin. It is also worth noting that the 30% of the vote that the actual CPC received in the 2004 election was significantly lower than the combined CA and PC share of the vote in 2000, which was just under 38%. It makes sense to assume then that not all Progressive Conservative voters would support a united Conservative party. Because of this, I also calculate seat totals for scenarios in which the Conservatives retain 80% and 70% of the PC vote. Because the Liberals are the closest party (excluding the Reform/CA) I assume that the remaining PC votes would have moved over to the Liberals. It is unlikely that PC voters would have jumped all the way across the political spectrum to the NDP, and the Quebec nationalist voters that the PCs won in the 1980s would have likely already moved to the Bloc Quebecois by the 1993 election.
It should be noted here that this is not an attempt to imagine a re-running of the 1993, 1997, and 2000 elections. There is no way to account for the way that an election campaign would have been different had the Liberals been running against a united Conservative party instead of a divided right. The way that parties approached the campaign, the strength of the united Conservative campaign, and the way that voters approached strategic voting all would have been different had the right been united. There is also no way to estimate the effects that having a united Conservative party in parliament between elections would have had on results. This is simply an attempt to understand the extent to which vote splitting on the right led to Liberal victories in the 1990s. It looks at whether uniting the Reform/CA and PC voters would have on its own been enough to beat the Liberals, not whether a united Conservative party would have been able to take votes from other parties in order to beat the the Liberals.
These hypothetical scenarios suggest that the split in the right did not play a large role in the Liberal victories of the 1990s. If the Conservatives had managed to keep all of the PC voters they would have managed to limit the Liberals to minority governments, but as noted above, that seems unlikely given the substantial numbers of Red Tories in the PC party. The graphs below show that if the Conservatives lost even 20% of PC voters to the Liberals, they would be looking at Liberal majorities once again. Only in 1997 is the result in the scenario where the Conservatives win 80% of the PC vote close, with the Liberals holding their majority by just two seats, but it is worth noting here that the Liberals were already extremely close to losing their majority when competing against the divided right. The 152 seats that Liberals win against the united right in the 80% retention scenario is only three seats fewer than the 155 that they won against the divided right. If the Conservatives lose 30% of the PC vote, the Liberals do as well as, if not better than, they did against the divided right.
While the split on the right did not seem to help the Liberals all that much, it did hurt the Conservatives. In each scenario presented, the united Conservatives take more seats than the divided parties combined. The problem for the united Conservatives face is two-fold. First, a small part of the increase in seats for the Conservatives is coming at the expense of the NDP and not the Liberals. This is useful for the Conservatives in the sense that it makes them stronger vis a vis the other opposition parties, but it does little to reduce the Liberal’s ability to form a government. Second, in the 80% and 70% retention scenarios the PC defections to the Liberals hand the Liberals a handful of Bloc and NDP seats that were otherwise close wins for one of the two non-conservative opposition parties. The graph below shows that Liberals pick up in between 2 and 9 seats in Quebec when they compete against a united Conservative party. This corresponds with a similar drop in seat share for the Bloc in every situation except for the 80% retention scenario in 1997. A united Conservative party would likely not have been able to take enough seats from the Liberals to both knock the Liberals down to a minority and compensate for the Liberal gains provided by defecting PC voters.
It is tempting to make the case that the Conservative victories have been in large part a result of the uniting of the right in Canadian politics. For all of the success that Harper has had, the right party vote in 2011 (just under 40%) was not much higher than the combined right party vote shares in 2000 (just under 38%) and in 1997 (just over 38%). Yet, if nothing else had changed, uniting the right in the 1990s would have been unlikely to lead to Conservative governments, let alone Conservative majorities. Having a united Conservative party certainly contributed to Harper’s success, but it was not on its own sufficient to produce Conservative governments. An understanding of how electoral politics has changed between the 1990s and 2000s requires looking at shifts on the left side of the political spectrum. The Liberal survival in the 1990s despite the loss of their Quebec stronghold cannot be fully understood without looking at the NDP and their collapse in that decade. This is something that I will examine in one of my future posts.
*Election results data that was used for these calculations can be found at Pundits’ Guide.
**Seat by seat calculations can be found here.