LibLead III

Finally, here is the third installment of three in my series on the Liberal leadership race.

In Part I, I guessed how well each leadership candidate would do as leader in a general election. To read it, click here.

In Part II, I explained that I thought each of the "big four" could realistically win the leadership race and I explained the scenarios as to how this would happen. To read it, click here.

Finally, here is Part III, which I described in my earlier post as:

A bit of the suspence for this one has been ruined as I have already said I think Kennedy will win. I will spell out why in a ballot-by-ballot guesstimation of the results.

So, here we go. For the first three ballots, I used a lot of data from the EKOS survey of delegates, which is the only opinion poll taken during this race that measures a relevant sample of people to determine the outcome of the leadership.

So, here we go.

First Ballot

For numbers, I used the results of the delegate selection meetings (excluding backfills, which would be too difficult to compensate for) and the ex-officio numbers from Wikipedia (which stood at: MI213, BR81, GK112, SD121, JV7, KD47, SB35, MHF3, Und/Neutral 476 when I prepared this).

Using the EKOS poll, I assume that 93% of Ignatieff delegates, 92% of Rae delegates, 94% of Kennedy delegates and 86% of Dion delegates will show up to the convention. I assumed that 60% of delegates for the bottom four and 75% of undecided delegates would show up. (This is elected delegates only).

For ex-officios, I assumed that 90% of those that had endorsed a candidate would show up and 75% of those who had not would show up.

For undecides, both elected, ex-officios who have not made endorsements and declared neutral delegates, I broke them down using the following logic. It is easy to endorse the frontrunner, so most who want to support him have done so. Thus, the undecides were broken down 10% for Ignatieff, 25% for Rae and 32.5% each for Kennedy and Dion.

This gives us the following results for the first ballot:

Ignatieff 1448 (31.2%)
Rae 977 (21.04%)
Kennedy 942 (20.31%)
Dion 843 (18.16%)
Dryden 152 (3.26%)
Volpe 131 (2.81%)
Brison 123 (2.66%)
Hall Findlay 26 (0.55%)

(Disclaimer: I personally think Martha Hall Findlay will do better than this, but couldn't think of a proper way to massage the numbers to make it so.)

Second Ballot

I assume that Dryden and Brison will choose to drop off and Hall Findlay will be forced to. Volpe I expect will remain on the ballot due to the way he has stubbornly approached the race and due to the rumours I have heard about the "deal" he wants in order to endorse a candidate.

Based on the EKOS poll, some pledged delegates to the big four do not want their candidate of choice to win. For Ignatieff, 11% want him to lose, for Rae 12%, for Kennedy 7% and for Dion 18%. For the second ballot, I have moved these people away from their pledges and allotted them to other candidates based on their second choice from the EKOS poll (undecideds excluded from the results).

For Volpe, I estimated 40% of his delegates would dessert him and assigned them to the second choices from the poll, undecides excluded.

For the three that dropped, leaving undecideds in the equation, I moved those who had made up their minds for a second choice to the big four. Then I moved all of their undecideds to who I thought they would endorse.

Based on everything I read in the press, I decided to have Dryden and Hall Findlay endorse Rae and take their undecided delegates (50% and 52% respectively) with them. For Brison, I had him endorse Kennedy - per my previous thoughts on this, click here - taking his 50% undecided in that direction.

This gives us the following results for the second ballot:

Ignatieff 1391 (30.08%)
Rae 1146 (24.79%)
Kennedy 1075 (23.26%)
Dion 920 (19.9%)
Volpe 91 (1.97%)

Note that on this stage, the only judgment calls I made were with endorsements. Other movement is based on the EKOS poll. Therefore, unless Ignatieff gets an endorsement from someone who drops off, and the EKOS findings hold, he is guaranteed to slip between ballot 1 and ballot 2 as the EKOS poll finds more people would bleed from him than would come to him from other campaigns. I found this very interesting.

Third Ballot

I begin this ballot by assuming that Volpe, who would be eliminated, "releases his delegates". Therefore I have accorded his delegates based on the EKOS findings of second choise, less undecideds. This means 60% go to Rae, 15% to Kennedy and 13% each to Ignatieff and Dion.

Then I assume that Ignatieff supporters, some of whom are with him because he is the winner, would give up. I estimate he would lose 15% of his delegates at this time, which is consistent with history of what happens when frontrunners stall. Of the delegates that leave him, I send 90% to Kennedy - as I suspect those leaving at this point would want to stop at all costs Rae and Dion who have attacked their candidate - and project that the remaining 10% will not vote or spoil their ballots.

Also, because Rae and Kennedy have both broken the psychological barrier of 1000 votes and Dion has grown, though not substantially and remains bellow both the 1000 vote and 20% barriers, 10% of his supporters move to other candidates. Of these, I move 10% to Ignatieff, 20% to Rae and 70% to Kennedy based on media reports of synergies and alliances between these two campaigns. I also projected that 10 Dion delegates would not vote.

This gives us the following results for the third ballot:

Rae 1219 (27.67%)
Ignatieff 1203 (27.31%)
Kennedy 1153 (26.18%)
Dion 830 (18.84%)

Fourth Ballot

At this point I would expect a lot of Ignatieff supporters to be frustrated and scared due to his not only having stalled but having also slipped to second place, I project a 30% bleed of his supporters. Of these, I project 70% go to Kennedy, for the reasons above, and 20% go to Rae, while 10% do not vote on the next ballot.

In Dion's case, he has dropped off, I suspect that Dion will endorse Kennedy based on the media speculation and other words heard on the street. However, I am not convinced he could deliver all of his delegates, especially when I already moved 7% of his delegates to Kennedy on the last ballot. I project 60% would follow Dion to Kennedy, while 30% would go to Rae and 10% to Ignatieff.

This gives us the following results for the fourth ballot:

Kennedy 1651 (40.83%)
Rae 1468 (36.29%)
Ignatieff 925 (22.88%)

Fifth Ballot

Ignatieff is eliminated here and per previous arguments, I suspect the lion's share of his support would move to Kennedy. I have 75% of his delegates moving to Kennedy, 20% to Rae and an additional 5% not voting.

This gives us the following results for the final ballot:

Kennedy 2345 (58.66%)
Rae 1653 (41.34%)