Monday, April 11, 2011

Conservatives and Liberals gain

Today's projection primarily incorporates two polls released over the last few days: Ipsos Reid from the weekend and Nanos Research from today. The last fully independent three-day Nanos poll has also been returned to the projection, along with the regional data from Forum Research that was provided to me.

In the end, though, things are still static and there are no seat changes in today's projection update.
That keeps the Conservatives at 153 seats, but they have also gained 0.4 points and now lead with an even 39%. The Liberals, unchanged at 72 seats, are up 0.2 points to 28%.

The New Democrats are down 0.1 point to 16.8% and are unchanged at 33 seats, while the Bloc Québécois is unchanged at 9.1% and 50 seats. The Greens, at 6%, are down 0.3 points.

There have been some more interesting shifts at the regional level, however.

In British Columbia, both the Conservatives and Liberals are gaining, displacing the NDP and the Greens.

In Ontario, the Liberals have remained stable at 34.6% but the Conservatives are up 0.3 points to 42.4%. The Liberals were getting close in a few ridings but this gives the Tories a little more breathing room. The NDP, however, are still dropping and are now projected to be at 15.7% support.

In Quebec, the Bloc is down 0.4 points to 37.1%, while the Liberals, Conservatives, and NDP all make small gains. If this continues, more than a few Bloc seats will switch over to the Liberals. Ahuntsic and Brossard - La Prairie are the ones most on the bubble.

And in Atlantic Canada the Liberals are in the lead with a big 0.8 point gain. A few ridings like West Nova and Saint John are looking promising for the Liberals.

At this point of the campaign, it appears that the head-to-head match-up between Stephen Harper and Michael Ignatieff is pushing Jack Layton aside, while in Quebec the Bloc is looking less dominant. It could be an interesting three weeks.


  1. BC is considered a battleground province but how many seats are actually up for grabs there?

  2. Good point Progressive Tory..

    In the 2008 election there were 18 seats that were within 15%

    The CPC won 3 close seats from the Liberal
    The Liberals won 5 close seats from the CPC.
    The NDP won 5 close seats from the CPC
    The CPC won 3 close seats from the NDP
    The NDP won 1 close seat from the Liberals
    The Liberals won 0 close seats from the NDP.

    Worst case
    CPC Loss of 6 seats
    NDP Loss of 5 seats
    Liberal Loss of 5 seats

    Best case

    CPC Gain 10
    Lib gain 3
    NDP gain 4

  3. Given the CPC "inefficient" vote pattern as laid out yesterday by Nick Nanos on CTV's Question Period I'm finding these projected numbers outside of the Prairies extremely suspect !!

  4. I am finding these numbers most attractive.

  5. Brossard is a Liberal seat already.

    Ahuntsic has lost Eleni Bakopanos as the Liberal candidate. Without her as the former MP and prominent member of the Greek community, I'm not sure Ahuntsic is attainable.

  6. Peter:

    You don't always get what you want!!

    The overall 41% in the Nanos polls is incredible inefficent ... giving the CPC only 145 seats.

    Eric's combined 39% gives only 153 seats.

    Chretien in 1993 41.2 % 177 seats out of 301
    Chretien in 1997 38.4 % 155 seats out of 301
    Chretien in 2000 40.9 % 172 seats out of 308

  7. Kind of sad when people can say (and be honest about it) that 41% of the vote getting you 47% of the seats is 'inefficient'. I really, really, really dislike FPTP.

  8. hosertohoosier11 April, 2011 12:04

    BC V of R, good point about vote efficiency. Although, Chretien was on the losing side of vote efficiency outside of Ontario. If you look at the 1997 results, for instance, and assume Ontario accounts for about 40% of the vote, you get the following support numbers and seat shares for each party:

    1997 Election without Ontario:
    Liberal: 31.5% (27.3% of seats)
    Reform: 19.8% (30.3% of seats)
    Bloc: 18.5% (22.2% of seats)
    PC: 19.14% (9.6% of seats)
    NDP: 11.45% (10.6%)

    The Liberal vote was only efficient in Ontario, due to the right wing vote split. We won't start getting to majorityville, until we start seeing 10-15 point wins in Ontario.

  9. All this speculation may be mute at this point. Check out CBC news story "Auditor draft report alleges Tories misspent G8 funds"
    Interesting tid bit just a day before the leader debates. Hard to say exactly how much impact this'll have on Tory fortunes, but it could be big. Remember the RCMP investigation of the Libs halfway into the 2006 campaign. It was devastating. This could be the straw that broke the camels, or reformers, back.

  10. I can't help but wonder what this recent heavy dose of Nanos polls has done to the projection. If we'd had just as many polls, but a more varied set of pollsters, would that have changed the projection outcome? After all, Nanos does poll the Greens very low and the Liberals very high. Your house effects correct for this, yes?

  11. "The overall 41% in the Nanos polls is incredible inefficent ... giving the CPC only 145 seats."

    Where do you see that? The single poll projection on the right is from the Nanos poll on mar 29 - that was 38.4, not 41.

  12. Let's hope the latest Harper G8 spending scandal finally sticks with voters.

    How many more scandals and ethically-questionable practices does one require in order to be turfed from office. The Liberals needed one... Harper is on, what, his 40th now??

    Canadians need to wake up and realize our Country has been taken for a ride down a poorly maintained dirt-road which ends at a cliff... Take a left people and get back on the highway to prosperity.

  13. Anonymous: It's not really fair to say "The Conservatives have had lots of scandals, the Liberals only had 1". The Liberal one was a BIG one, involving envelopes of cash going from the government to the Liberal party. For all the Conservative scandals, they're all MUCH smaller than that.

    Eric, I'm curious if you could do a more recent run-down at the ridings that are projected to be potentially in play, and what sort of swing would be required to move substantial numbers of them? I'm particularly interested to know if there are any watersheds... lots of ridings where a similar sized swing would tip the riding.

  14. @John Northey,

    I think it's less a situation where FPTP is the problem and more a case where multi-party systems are the problem.

    From everything I've read / seen, in a multi-party system, you generally end up with one of the two following situations:
    - King-makers that have a disproportionate influence over parliament versus their popularity. This is because parliament becomes very fractured and in order for things to pass the smaller parties get a significantly stronger influence than they would otherwise (a parallel is how swing ridings have more influence on policy than other ridings in election campaigns).
    - Leading parties that end up with more clout in government than their popular vote would indicate (the current situation).

    I'm not claiming two party-systems are better. There is no perfect system out there and everything has its drawbacks.

    I tend to like the way Canada's system works pretty well (and I've said this over the last 20 years or so, so it's not tied to the success of my party). We get an influx of new ideas (i.e. reform party and green party - who does influence the dialogue moreso than a similarly popular party would in the US for example). At the same time, we have enough concentration of power in the party with the most votes that there is a focused government rather than having a government that tries to be all things to all people.

    My biggest complaint is the polarization that has taken place at the moment. I find it strange that the two most similar, and most popular, parties are two that refuse to work together - each in hopes of having their own majority eventually. It's odd that generally each opposition party came very close to supporting (or did support) each of the Conservative budgets, including the one that failed, but come campaign time, the one party that no one includes in discussions of potential coalitions is the Conservatives, the most popular party (seats or votes).

  15. @Anonymous ...

    The Liberals had much more than one scandal. There was Shawinigate, the choking, HRDC, the Somali cover-up, APEC, etc.

    Conservatives have scandals and try to cover them up. Liberals had scandals and tried to cover them up. At the end of the day, how they govern is less important to me than the bills / laws they pass as that's the only place I see a difference.

  16. I'm a partisan Liberal who lives on Vancouver Island, so I don't write this with any partisan intent to sway your predictions... but I have to ask you where in god's name you got your projection that the Liberals will win Esquimalt - Juan de Fuca by 39.9%?!

    I don't know a single local observer who views the race with any sort of that lead for the Liberals... If they do win, it will be by a very close margin, with the advantage still going to the Conservatives.

    Former MP Keith Martin was widely regarded as a "maverick" MP who became a Liberal after sitting with the Reform and Alliance parties. He squeaked through an 80 vote win against Troy Desouza (who's running again and has spent every waking second of his life campaigning since the last election) partially due to NDP voters strategically voting to save Martin at the 11th hour (this is according to poll by poll analysis from a campaign worker--can't confirm if it's true).

    Before Keith Martin went Liberal, there was little actual partisan Liberal support in the riding--it used to be heavily a Conservative versus NDP affair. Now that Keith Martin's gone, it'll be hard to assume that the effect will continue. Many seem to believe that Garrison from the NDP is now the "ABC" candidate. I'm not entirely correct that's accurate, but I know this time around the NDP will NOT strategically vote for the Liberals like they did for Keith Martin.

    So unless you have local riding polling data that can counter this, I think you are applying a broad brush of provincial polling data to this riding that is currently facing very complicated local factors--none of them indicating that the direction of your prediction is right. Your prediction of Esquimalt-Juan de Fuca is important to get right, because this riding is widely regarded as one of the ridings that will determine whether or not Harper can pickup a majority.

    Pundits Guide wrote a great analysis of this riding here:

    Even the subjective election prediction project widely casts doubt that the Liberals are in any real 'lead' here to win:

    And again another local blog making the same reflection that your numbers here are completely skewed and inaccurate for Esquimalt - Juan de Fuca:

  17. Ira

    Where do you see that? The single poll projection on the right is from the Nanos poll on mar 29 - that was 38.4, not 41.

    You are correct. But still 38.4 for 145 seats is very inefficient compared to Chretien's 38.4 giving his winning legacy stong majority 155 seats out of 301.

  18. "Esquimalt - Juan de Fuca is not even close to being accurate"

    Why does your "name" link to PG?

    The model requires a minimum of uniformity, otherwise it would just a be subjective amalgamation of my guesses.

  19. Eric,

    Could you subtract the star candidate factor for Keith Martin now that he's gone, just as you added it for other star candidates?

  20. The Liberals are already penalized by not having an incumbent in the riding - subtracting the "star candidate" factor would have about the same result.

  21. What would the numbers be if you subtract the 'star candidate' factor in addition to the non-incumbent factor?

  22. Esquimalt - Juan de Fuca is not even close to being accurate11 April, 2011 15:20

    I'm not associated with pundits guide, I was unsuccessfully trying to link to the post they had about Esquimalt - Juan de Fuca, where they brought up a lot of the same points I did.

    But I'm still trying to understand what polling data you are relying on for your Esquimalt - Juan de Fuca projections? If I was taking bets on this riding, I would definitely raise my projections for NDP performance (you have their performance declining 2% from last time in this riding... it's certainly going to be higher this time around) and lower projections for Liberal performance.

    I'm not ready to say either the Liberals or NDP are in the lead against eachother, but I do believe that right now the Tories have at least a slight lead on both that will propel them to take Esquimalt-Juan de Fuca.

    On a personal note, I honestly think Keith Martin didn't run again after his close win because he didn't think he could win it again against Troy Desouza... but that's just speculation of course.

  23. Anonymous 15:10,

    Even then the Liberals would have a slight edge. It speaks to the amount of growth the Liberals have had in BC, but also how Esquimalt-Juan-de-Fuca is an exceptional riding.

    I'm afraid I can't just lop off 10 or 15 points from the Liberals arbitrarily.


    I'm not relying on any riding-level polling data for Esquimalt-Juan-de-Fuca. I invite you to read about my methodologies, which are fully explained on the site.

    The link to these explanations is just under the LE DEVOIR image, but here is a direct link:

  24. Writing about Jeanne--Le Ber's projections again, there is a 0.7% "other" column, but according to the elections canada website the only other registered candidate is a Marxist-Leninist, and the last time they ran a candidate here they got 0.32% Is 0.7% just the leftovers because of your mathematical prediction model, or do we actually expect the Marxists to double their vote in this election?

  25. Anonymous 15:52,

    The projection hasn't been updated for the Other candidates yet. I believe nominations close tomorrow, so look for those numbers to be updated this week.

  26. jbailin, Mike:
    I get your points, but I think they miss the point. The debates start tommorrow. Most voters will start paying attention to the election about tommorrow. It doesn't matter if Liberals had scandals or not. Electors for the most part have short memories. This will hurt the CPC. The timing is too critical.
    Oh yeah, and we're talking $50 million here. That's a lot more than the amount Libs are alleged to have skimmed from adscam.

  27. @pinkobme ...
    "Oh yeah, and we're talking $50 million here."

    That's a rounding error for HRDC, which was at least 20 times as large.


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.