Wednesday, March 13, 2013

Garneau out, Trudeau definitely in

A bombshell landed in the Liberal leadership campaign today when Marc Garneau announced he was dropping out and throwing his support behind Justin Trudeau. But the outcome of the race was hardly in doubt before Garneau's withdrawal, and apparently the former astronaut agreed.

I haven't posted the endorsement rankings on the front page since they were published for the first time on January 11. Instead, I've just been updating the 2013 LPC endorsement rankings page each Friday with a little analysis. It simply hasn't been very interesting: only Trudeau has gotten anything more than a handful of endorsements, and his total share of the endorsement points has never dropped below two-thirds. But with Garneau's departure being the biggest news of the race so far, and that departure including the news of a survey of Liberal members and supporters, it seems like a good time to take a look at the numbers.
The only endorsement added to the rankings since last week's update is the tepid one from Marc Garneau. Trudeau gets all five of Garneau's points as he is a two-term MP, while he gets half of the points Garneau had gather from other endorsers (unless those endorsers head elsewhere in the coming weeks).

It pushes Trudeau up to 94.6% of the endorsement points, an almost comically high result. Joyce Murray now moves into second with 3.8%, while no other candidate has more than 1% of the endorsement points. Nevertheless, the top four in the rankings are likely who we will see in the top four on the first ballot of the leadership vote.
If we weigh those points by province, as the Liberal Party will be doing by giving each riding equal weighting in the vote, Trudeau still gets 91%, followed by Murray at 6.5%.

Compared to the last time the endorsement rankings were on the front page, Trudeau has increased his total share by eight points and his weighted share by 19. Murray's total share has not changed, but her weighted share has fallen by six points.

So, it looks like a Justin Trudeau landslide. The vote will almost certainly be closer than the rankings imply as each of the campaigns do have their own block of voters - and many of them will probably be more motivated to vote than many of Trudeau's supporters (you'd have to be pretty committed to support one of the other candidates in this race).

Does that make it a coronation? I find that description very odd. The race started with nine candidates, and there are still seven in the running. Marc Garneau, Joyce Murray, and Martha Hall Findlay are not wilting flowers, and have tried to push Trudeau. They had every intention of winning the race when they launched their bids. The fact of the matter is that none of them were able to gather as much support as Trudeau was. That does not make it a coronation - it would only be a coronation if all of the plausible candidates stepped aside, as occurred when Paul Martin and Michael Ignatieff were selected. Trudeau is going to win this easily simply because the majority of Liberal members and supporters see him as the best candidate. Just because it isn't close doesn't mean it is a coronation - it just means that the other contestants were not seen as a better option.

That is what Marc Garneau concluded when he saw the results of an internal party survey his campaign conducted.
Marc Garneau made an effort to call this a survey and not a poll, as it does not appear to have been your standard scientific poll with a random sample. Instead, 50,000 Liberal supporters and members were invited to take part, and some 6,000 did. Who those 6,000 are is not known - they could have disproportionately come from one camp or another, or from one region of the country, or one demographic group, etc. Perhaps the Garneau campaign has more information, but we don't have anything else to go on but the numbers he reported.

Nevertheless, the results are pretty conclusive. With 72% in the survey, Trudeau is so far ahead that it puts him safely outside the potential errors and biases that the survey's methodology could inject into the sample. If Trudeau was closer to 50% then there might be some question - but this huge number is pretty definitive.

Where the other candidates stand is, on the other hand, hardly definitive. Garneau's poll gave him 15%, Joyce Murray 7%, and Martha Hall Findlay 5% (it appears no other candidates were included in the survey). Because of the potential problems of the survey, any of the three could actually be second. In fact, Murray's campaign was quick to come out and say that they were second, and that Garneau was much further behind. We'll have to see if any of them back this up with their own numbers.

Sound familiar? It should - during the NDP leadership campaign in 2012 the Dewar and Mulcair campaigns each released competing polls that placed Brian Topp in third or tied for fourth. The Topp campaign responded that their own numbers showed Topp to be much higher, and in the end they turned out to be right (Topp finished second on the first ballot). Peggy Nash, who was placed as the runner-up to Thomas Mulcair in both the Dewar and Mulcair polls, ended up placing a rather distant fourth.

Who is the Topp in this case? Is it Garneau, who the other campaigns claim is not second, or is it Murray or Hall Findlay, who were placed well behind in Garneau's poll? I suppose we'll never find out where Garneau would have finished, but we will find out who places second if Murray and Hall Findlay stay in the race.

They probably should. Joyce Murray's campaign is the only one apart from Trudeau's that could claim to be holding some momentum. She's gotten a lot of endorsements from outside the party, and a few from within the party itself, and could surprise in the same way that Nathan Cullen did as the co-operation candidate (though Cullen's surprising result also had to do with his strong performances in the debates). But because of her very controversial plan, she does have a bit of a cap in terms of her potential support. It is difficult to imagine a scenario in which she could win, but a very good showing could make the party consider her plan (though probably not).

Martha Hall Findlay can capitalize on Garneau's departure by being the primary non-Trudeau option (Martin Cauchon might challenge that assertion). The two might have split the non-Trudeau vote had they both remained in the race. Instead, Hall Findlay might be able to corral enough of Garneau's vote to finish second ahead of Murray. That would be important for her own political future.

It all might be a little academic, though. If Garneau's survey is even slightly accurate, Trudeau has a very good shot at winning on the first ballot. I would have said that, against seven other candidates including three with good bases of support, Trudeau would have had difficulty getting over the 50% mark on the first ballot. But now that the list has shrunk to six opponents, only two of whom are likely to break double digits, it seems likely that he can get enough to win on the first ballot.

But a landslide first ballot win is probably not in the cards. If Karen McCrimmon, Deborah Coyne, and David Bertschi all hold on until the vote, they can probably still pull together some 5% to 10% of the first choice ballots (Martin Singh and Niki Ashton together managed almost 12% on the NDP's first ballot). Cauchon might be good for another 5% to 10%. That leaves 80% to 90% of the ballots still available to the other three contestants, meaning a best case scenario requires Murray and Hall Findlay to combine for more than 30% support, or 40% in a worst case scenario, in order to prevent a first ballot victory for Trudeau. That might be difficult to achieve. It isn't impossible, but Murray and Hall Findlay did not manage that in Garneau's survey, even if we give them all of his support.

Could Trudeau have been more greatly tested in this leadership race? Certainly - and it might have done him some good. But even his critics have to admit that he has run a very competent and well-organized campaign that would have been difficult for any candidate to have beaten. Now it's Harper and Mulcair's turn to try.

19 comments:

  1. You tend to lose credibility as a party when one of your candidates endorses someone they have previously attacked so fiercely.

    ReplyDelete
    Replies
    1. No you don't. Garneau did his best to test Trudeau in an attempt to garner support. It was a good try but it didn't work. I realize the theme of the other two major parties' supporters will be this is a coronation and the Libs are destined for the dust bin. Unfortunately for the NDP and CPC that is not the case. They almost pulled it off in 2011, but failed. That was their (the CPC and NDP) best chance and they blew it. That train has now pulled out of the station. If he's elected leader, Trudeau, who is popular in all regions of this country, will make the next election very difficult for the current government and opposition parties.

      Delete
    2. Well said, pinkobme!

      Delete
    3. I can't actually think of any instance where the conduct of a candidate who didn't win a leadership seriously affecting a Party....

      Delete
    4. I'm almost certain the Garneau survey is results from voter identification calling (therefore it would have identified the source). You would never get such a high response rate on an IPR and it would be really pointless and expensive for a failing leadership campaign to conduct a live poll with such a large sample size.

      Delete
  2. I would be careful about taking that survey to the bank depending on how it was conducted. Rumour has it candidate support varies sharply based on how members/supporters were recruited (ie by phone, in person or online), and that the support of some campaigns is highly concentrated in a relatively small number of ridings.

    ReplyDelete
  3. Hopefully this isn't a double post on my part, but don't take that survey to the bank. Rumour has it support for each candidate varies quite wildly based on how they were recruited. If this survey was done by phone, it may not have captured many supporters from candidates like MHF and Murray who focused on signing up people online. Based on what people are saying about the distribution of membership numbers, it looks like there were big sign-ups in a small number of ridings that are skewing the picture a lot too.

    ReplyDelete
    Replies
    1. I heard that they took the riding distribution into account.

      Delete
  4. "Garneau did his best to test Trudeau"
    OR did his best to be his sparring partner to "toughen him up" for the battles that lie ahead? Garneau bent his knee a week ago. No surprise with his withdrawal and endorsement.

    ReplyDelete
  5. Cabinet Minister post confirmed!

    ReplyDelete
  6. Ah, apparently these were Robocalls that Garneau made over the last couple of weeks. It would be expected to skew Trudeau, as Trudeau relied on phone banks for sign-ups. How much of a skew is something I don't know.

    ReplyDelete
  7. Frankly I believe Garneau got out because it looked increasingly like Murray's insurgent campaign might vault her into second place.

    The Liberals remind me of the Black Death skit by Monty Python where the body being carried out to the cart keeps saying "I'm not dead yet".

    It is like 1) Liberals do we want a party - answer Yes of course. 2) Why do we want to have a party? Hmmm trick question I wasn't expecting that, I'll get back to you.

    ReplyDelete
  8. I am not buying this. If Trudeau had such an insurmountable lead as the media would lead you to believe, why Trudeau campaign was in a panic mode, demanding extension for the deadline? If his victory was an assured thing, why create all this negative publicity for the party? I did my math from the leaked numbers, and it appeared that less than half of 90,000 registered voters were from Trudeau camp - so he wouldn't win in the first ballot, hence the panic.

    Again, I might buy Garneau's reason for quitting as "mathematically impossible to win", but why endorsing the man you had been critisizing heavily just a few days ago? The timing of the Garneau departure, and this unnecessary endorsement again suggest that the party establishment is not certain at all that their man (Trudeau) will win.

    And then this unscientific poll. Eric, I thought you'd be a bit more critical about it - this is your domain after all. Garneau didn't say how he chose the numbers to call - right? He called 50,000, with 90,000 registered, and almost 300,000 signed up. Did he call anyone signed up by Murrey? Perhaps he didn't, as he'd find almost zero support for himself in that group, and he needed to estimate his chances against Trudeau in the groups were he thought he was doing well (party membership + supporters of Martha etc.) - so why waste your money? Joyce mentioned "tens of thousands signed up", and I can't imagine that translating into 7% of support, with 90,000 registered. I'd think this should be more like 30%.

    ReplyDelete
    Replies
    1. I pointed out that this wasn't a scientific poll and laid out the potential problems. If Garneau had presented it as one, I would have been more critical. We often see these kind of surveys during leadership campaigns.

      Trudeau wasn't the only campaign to ask for an extension, so I wouldn't read too much into it. They probably just want to make sure the number is as high as possible. Having only 1/3rd register is worse publicity than asking for an extra week and bumping that up to 1/2 or more.

      Also, we don't know how the original members of the LPC would vote - if they break down similarly to how other Liberals do, it doesn't matter much if Trudeau's campaign is having trouble registering their supporters.

      You are making some big assumptions with Murray - she said tens of thousands, which could mean as low as 20,000, even if they weren't exaggerating. If she is also having the same issues registering supporters, that might mean only 7,000 or so of the people she signed up have registered. When you consider that, 7% is not implausible.

      In the end, it doesn't matter much. The odds that the survey would be that wrong are quite low, unless Garneau ran a completely incompetent survey and then decided he would withdraw because of it. Considering his background, that seems unlikely.

      Delete
    2. Murray opposed the extension and MHF neither supported nor opposed it (the other campaigns were all in favour). That suggests to me that Murray and MHF aren't having the same troubles registering their supporters as Trudeau.

      Garneau's survey also apparently prompted in the following order every time: Garneau, Trudeau, MHF, Murray (did not mention any others. Unlike a traditional poll this wasn't rotated. I wonder if a polling firm would know how much one could expect that to skew things? I'd think it would favour Garneau and maybe Trudeau a little bit.

      My understanding was that Garneau felt the regional distribution was not in his favour based on fundraising and total signups information, and not necessarily this survey, though I'm not 100% sure on that.

      Delete
  9. And now the Elections Commissioner is calling for charges over the Robocall scandal !!

    ReplyDelete
  10. Perhaps I missed it, however I noticed in your endorsement ranking that David Suzuki's public endorsement of Murray is not counted in her endorsement points. While not a politician, how would you factor in 'public figures' into endorsement points?

    ReplyDelete
    Replies
    1. The endorsement rankings are supposed to be about "establishment" support, so I only include elected or formerly elected party members.

      I wouldn't know how to quantify an endorsement like Suzuki's.

      Delete
    2. If my memory serves me correctly Dr. Suzuki was destined to become Joe Clark's minister of the environment before the government fell.

      Delete

COMMENT MODERATION POLICY - Please be respectful when commenting. If choosing to remain anonymous, please sign your comment with some sort of pseudonym to avoid confusion. Please do not use any derogatory terms for fellow commenters, parties, or politicians. Inflammatory and overly partisan comments will not be posted. PLEASE KEEP DISCUSSION ON TOPIC.