The Conservative policy convention being held this week in Vancouver will help determine the future of the party. But how much of the party's divided past is still a factor, more than 12 years after the Canadian Alliance and Progressive Conservatives united the right as the Conservative Party of Canada?
Compared to the Liberals and New Democrats, the Conservative support base is split more evenly between its centrist supporters and those on the edges of the political spectrum.
Polling data provided by Abacus Data suggests that 45 per cent of Conservative voters self-identify as being centrist, compared to 45 per cent who say they are either on the centre-right or right.
That is a greater division than either the Liberals or the New Democrats have between their centrist and left-wing support bases. Fully 54 per cent of Liberal voters and 50 per cent of New Democrats consider themselves to be in the centre, compared to 31 and 39 per cent, respectively, who say they are on the left or centre-left.
- The Pollcast: The state of the Conservative leadership race
- Analysis: How Conservative and NDP leadership contenders stack up on the money
This suggests that, of the three largest federal parties, the Conservative party has less of a consensus among its supporters concerning where the party lies — or should lie — on the political spectrum.
One member, one vote?
The Conservatives will elect a new leader in May 2017, and a leadership race is officially underway, although with just three candidates declared so far.
One of the issues frequently debated at Conservative conventions relates to the rules for electing party leaders. Unlike the NDP, which gives every party member's vote an equal weight, the Conservatives and Liberals use a different system that gives each riding in the country an equal weight.
That means that whether a riding has 1,000 members or 100, it is worth the same number of "points" in a leadership contest.
This has led to heated debates on the convention floor between the old PC and Reform wings of the party. Peter MacKay, the last leader of the PCs and a potential contender to replace Stephen Harper, has been the most vocal defender of the rules in place.
These rules ensure that parts of the country without large membership rolls, such as Quebec and Atlantic Canada (traditional bases of the Progressive Conservatives), are not swamped by the regions of the country with more members.
An amendment proposal to change the rules to a one-member, one-vote system will be considered at the convention again this year and may come to a vote.
Not all votes are equal
Those who believe the current rule system disadvantages some members may have the numbers to back up that claim. An analysis of financial contributions to the party suggests that the party's membership is disproportionately weighted towards Western Canada, and is far less represented in Quebec.
While it is not the same as having access to a complete party membership list, past donations made to each electoral district association (EDA) in the country acts as a decent proxy for determining the relative weight each region of the country carries within the Conservative Party.
According to Elections Canada data, between 2012 and 2014 (full details for 2015 are not yet available) 46.3 per cent of individual contributions were made to Conservative EDAs in Ontario. Another 41.6 per cent were made to Conservative EDAs in British Columbia, Alberta, Saskatchewan, and Manitoba.
Only 5.5 per cent of individual contributions were made to EDAs in Atlantic Canada, while just 5.4 per cent were made to Quebec.
If we can infer from this that roughly half of Conservative members are located in Ontario and about two-in-five are in Western Canada, we can see just how much the current voting system favours Conservative members in Eastern Canada.
With 78 ridings, Quebec will represent 23.1 per cent of the "points" awarded in the Conservative leadership race. Atlantic Canada has 9.5 per cent of the country's ridings, and will get 9.5 per cent of the points. Ontario and Western Canada have 35.8 and 30.8 per cent, respectively — far short of the membership weights suggested by their donations in those regions.
Current system could favour PC candidate
Without these rules in place, a candidate hailing from the Progressive Conservative tradition in the party might find it very difficult to win a leadership contest.
In the three federal elections held between 1993 and 2000, when the right was divided, the Progressive Conservatives took about six times as many votes as the Reform Party or Canadian Alliance did in Quebec and Atlantic Canada. In Western Canada, however, Reform or Canadian Alliance candidates took about four times as many votes as their PC rivals. They were also more successful in Ontario.
Let's apply those ratios to a hypothetical contest between a candidate from the PC wing and a candidate from the Reform wing, using the estimated breakdown of the Conservative Party's membership today. In such a scenario using a one-member-one-vote system, all else being equal, the PC candidate would take under 40 per cent of the vote.
In the current weighted system, however, the PC and Reform candidates would split the vote evenly between them — thanks to the boost the PC candidate would get from the sparsely-membered ridings in Quebec and Atlantic Canada.
For proponents of the principles of the 2003 merger, that might prove the system's fairness. It means that candidates hailing from either wing of the party have an equal chance of winning, with the title going to the candidate who is best able to bridge the gap between the two factions.
Laudable, perhaps, unless one believes that every party member's vote should be created equal.
Party members may not want to open old wounds this weekend. But if the policy convention avoids the divisive issue, will the results of the leadership campaign see it reignite again?
- The Pollcast: Subscribe to Éric Grenier's podcast
- Poll reveals which historical figures Canadians want as the woman on their money
The poll by Abacus Data was conducted between May 17 and 20, 2016, interviewing 2,000 adult Canadians via the Internet. A probabilistic sample of this size would yield a margin of error of +/- 2.2 per cent, 19 times out of 20.