How the Conservative leadership vote could be won, ballot by ballot

Maxime Bernier is the favourite to finish atop the first ballot for the Conservative leadership, but an analysis of how the vote could play outsuggests that it could take all 13 ballots before a winner is finally declared.

And while Bernier remains the favourite to win it all, his path to the leadership is far from certain.

The next leader of the Conservative Partywill be decided by a preferential vote, in which members can rank up to10 of the 14 candidates on the ballot.

• Almost half of Conservative members have voted
• Counting points, not votes, in Conservative leadership race

Points will be awarded according to the share of votes candidates receivein each of Canada's 338 ridings. Unless one candidate has a majority of points, the candidate with the fewest points nationwide is eliminatedand their votes redistributed to their supporters' next preference. This process continues until one candidate reachesa majority.

It is possible to estimate the first ballot result based on available data (fundraising, polls and endorsements). The Conservative Leadership Index, the latest edition of which can be found at the bottom of this article, does just that.

Where the vote will go after thefirst ballot is more difficult to figure, but it is possible to get anindication of voters' preferences from fundraising data.

A significant proportion of donors contributed funds to more than one candidate, suggesting that they are likely to rank these candidates highly on their ballot and presumably like-minded voters will do the same.

Scheer, Bernier with second choice potential

An analysis of fundraising between Jan. 1 and Mar. 31 shows that Bernierand Scheer share themost donors with the most other candidates.

Berniertopped the list among donors to Michael Chong, Kellie Leitch, Erin O'Toole and Lisa Raitt. Scheer was the top candidate among donors shared with Bernier, Chris Alexander and Rick Peterson, while O'Toole was first amongst Andrew Saxton's donors.

Bernier and Scheer ranked highly among other candidates as well.

The most significant overlapwas between Trost and Lemieux, two social conservatives. Fully 41 per cent of Trost's shared donors contributed to Lemieux's campaign, while 28 per cent of Lemieux'sshared donors contributed to Trost's.

It is possible that some of these shared contributors paid fees for fundraising events attended by multiple candidates, though the data does not suggest that those donorsrepresent a significant proportion of the whole.

Less than 15 per cent of donors who contributed to the campaigns of Bernier, Chong and Leitchgave money to another candidate. Contributors to Blaneyor Trost were the most likely to have given to someone else, representing 42 and 37 per cent of these two candidate's donors, respectively.

Going ballot by ballot

Using this data to estimate the preferential ballotand potential dropped ballots (if voters only mark their first choice), as well as combining it with the Conservative Leadership Index and an analysis of each candidate's distribution of support by riding, it is possible to simulate how the vote could break down ballot by ballot.

This is not a prediction of the outcome, but rather an illustration of what implicationsthe preferential ballot could have on deciding the winner.

The first few ballots do not have a big impact. Combined, Deepak Obhrai, Peterson, Alexander and Saxtonare estimated to take just 4.6 per cent of the points. By the sixth ballot (assuming Kevin O'Leary, whose name will be on the ballot, is among the first to drop off) no remaining candidate could be more than apoint further ahead.

The seventh ballot, which in this scenario redistributesBlaney's 2.9 points, primarily boosts Lemieux and Leitch. If Blaney performs better in Quebec, however, he could find himself surviving beyond the seventhballot.

On the eighth ballot, Trost sends most of hispointsto Lemieuxand a significant proportion to Scheer. This pushes Lemieux past Chong in the rankings and narrows the margin between Scheer and Bernier. Raittis eliminated.

On the ninth ballot, Bernier, Scheer and O'Toole receive a bump from Raitt but Chong does not getenough of a boost to push him past Lemieux and save him from elimination.

On the 10th ballot, Bernier and O'Toole gain the most from Chong'souster, Lemieux and Leitch the least. Along with Raitt'selimination this could be anopportunity for O'Toole to move ahead of Scheer. But both are still well behind Bernier who, at32 points, remainsfar from a majority.

Lemieux drops off and on the 11th ballot a big share of his support goes to Scheer, allowing him to further close the gap on Bernier and widen the margin between him and O'Toole. It is possible that Lemieux could instead give a greater boostto Bernier, who has emphasized during his campaign that under his leadership social conservatives would be free to express their views.

With the elimination of Leitchon the 12th ballot, Bernierstill leads with 38.4 points, but Scheer (33.6 points) and O'Toole (28 points)are not far behind.

Leitch's voters could play a decisive role. They are a large enough group that they could decide whobetween Scheer and O'Toole goes head-to-head against Bernier. If a big enough number of them instead opt forBernier, his win would be all but assured.

In this simulation it is O'Toole who drops off the 13th and final ballot. The fundraising data suggests that Bernier and Scheer split his voters, giving Bernier the win with 52.9 points to 47.1 points for Scheer. Bernier'swide advantage in Quebec, where he wins 74 per cent of the points, gives him the victory.

Endless possibilities

But this simulation presentsa close enough margin and alternative turning points for it to easily go Scheer's way(or O'Toole's,if he is the last man standing with Bernier). In order to win in this simulation, Scheer would need to take 60 per cent of O'Toole'spoints. That is achievable. Bernier would need a bigger margin by the penultimate ballot to hold off a consolidation of Scheer's and O'Toole's voters.

It would not take much else to change the outcome. Greater support for and from Lemieux and Trost could put Scheer in a better position by the final ballots. O'Toole could over-achieve with significantsupport from Raitt's and Chong's voters.With a better first ballot result or more second choice preferences throughout the count, Bernier could put himself out of reach earlier on maybe even winning before the 13th ballot.

Bernierhasraised the most money during this campaign and is polling ahead of his rivals. He is the favourite going into Saturday's vote. But that shouldn't dampen any of the suspense before the winner is announced.

The index is based onfour different metrics: endorsements, fundraising, contributors and polls. In tests on 14 recent federal and provincial leadership races in which all party members could vote, the index has replicated the first ballot resultswith a medianerror of +/- 2.2pointsper candidate.

A more detailed explanation of the index's methodology can be found here.