It's all over but the voting

With apologies to the Toronto Star, I think their new poll all but guarantees that my candidate of choice, Gerard Kennedy, is all but guaranteed to win the Liberal leadership on December 2.

The Star/EKOS poll is a survey of almost a quarter of Liberal delegates to the Montreal convention. The Star bills it as good news for Stéphane Dion, but I just don't see it.

I think most people who are watching this race closely, it is becoming more and more evident that whichever of Mr. Dion and Mr. Kennedy finishes third on the ballot with four candidates will win the leadership race. Why? Their combined vote totals put them in first place and their campaigns have almost publicly admitted an alliance which would see one go to the other.

So why do I see Kennedy as the candidate pulling in third?

First, the important fact to cite is that Kennedy currently stands at 17.3%, while Dion stands at 16.0%... that is a significant gap for Dion to make up over the course of one or two ballots until one of them is eliminated. Thus, I have always thought that the odds are that it would be Kennedy, not Dion, that would be in the showdown with Ignatieff and Rae.

So what does the poll tell us? Dion has the greatest growth potential of any candidate, therefore, the Star argues, he will surge and win. However, growth potential is largely irrelevant until one of the "big four" drops off and a large chunk of delegates become avaiable. In order for Mr. Dion, or another candidate for that matter, to exploit major growth potential, they must first hold in third or higher until the weakest of the big four falls.

According to this poll, in my view, it is unlikely that Mr. Dion will be anything but the weakest of the big four.

A few selected quotes from the Star, emphasis added by yours truly:

Asked if there is "any candidate you would definitely NOT like to see win," ... Gerard Kennedy, who resigned as Ontario education minister last spring, scored the lowest at 2 per cent.

EKOS found 21 per cent of delegates prefer Dion as their second choice, ahead of Rae and Kennedy. Trailing was Ignatieff, with 8 per cent. (It is unclear where Kennedy stands here but it is less than 21% but higher than 8%, presumably significantly higher)

The undecided vote is also a big unknown. Asked for their second choice, 38 per cent of delegates declined to answer.

(Kennedy's) delegates are "fiercely loyal, determined to come to the convention and convinced he is the man to beat the Conservatives in the next election."

In contrast, Dion had the lowest delegate attendance intentions, with 13 per cent of his delegates saying they are unlikely to be able to travel to Montreal.

Without tapping in to the support of Ignatieff, Rae or Kennedy's supporters, Dion has to make up 1.3% on Kennedy while 13% of his supporters (2.1% of convention delegates) are unlikely to attend. So, with only a small fraction of delegates opening up, Dion needs a gain in the neighbourhood of 3.4% to pull ahead of Kennedy. This is very unlikely.

Thus, it would seem, the inevitable result would be for Dion to drop and go to Kennedy, catapulting him into a strong second or a weak first. It also seems likely that whoever is in third at this point, either Bob Rae or, less likely, Michael Ignatieff, they are likely to go to Kennedy due to the growing animosity between the two candidates and their supporters.

Thus, ladies and gentlemen, it's all over. Kennedy wins.