Monday, April 4, 2011

Small gains by Conservatives and NDP

With new polls from EKOS and Léger Marketing over the weekend, we now have some other polling firms to add to the model in addition to the daily numbers from Nanos Research. A full poll summary will follow later today.

The new projection involves quite a few seat changes since Friday, but in the end the net result is a gain of one seat by the Conservatives and one by the New Democrats, both at the expense of the Liberals. But in all, five seats have changed hands.
The Conservatives have picked up 0.2 points nationally and now lead with 38.8% support and a projected 152 seats. The Liberals have dropped 0.1 point and are down to 27.1%, putting the gap between the two parties at 11.7 points (it was 11.4 in 2008). The Liberals have slipped to 71 seats in the projection.

The New Democrats are up 0.3 points to 16.9% and 34 seats. Will they get to 17% tomorrow?

The Bloc Québécois is down 0.2 points nationally to 9.4% but is unchanged at 51 seats, while the Greens are also down 0.2 points to 6.5%. Note that is lower than their level of support in 2008.

Regionally, only the NDP had a good weekend. They were steady in the Prairies and Ontario, and made decent-sized gains in Alberta, Quebec, and Atlantic Canada. With the possible exception of Quebec, however, those provinces are not growth regions for the NDP.

The Conservatives made a gain of one point in British Columbia, but also lost about a point in the Prairies. A drop of 0.4 points in Ontario is also problematic, considering the 0.5-point gain of the Liberals there.

But aside from that and a 0.7-point jump in the Prairies, this update has little good news for the Liberals. They dropped big in British Columbia and Atlantic Canada, and dipped below 20% in Quebec.

Four of the five seat changes took place in Ontario. Ajax - Pickering went from the Liberals to the Conservatives, and now Christopher Alexander is projected to be in the lead. It is the same story in Brampton West and Brampton - Springdale, where Kyle Seeback and Parm Gill are favoured.

This might seem odd considering that the Liberals closed the gap in Ontario by 0.9 points. I won't try to rationalize this change with some theory about voting behaviour. This is just how the model works. The incumbency effect is partly dependent on whether a party is gaining or losing in a province. In this update, the Liberals have gone from marginally losing support to marginally gaining support compared to 2008 in Ontario. That means the incumbency effect for MPs like Ruby Dhalla and Mark Holland has flipped, putting them at a disadvantage. The incumbency bonus is far greater when a party is losing support than when it is gaining, as my own analysis has shown.

One piece of silver lining for the Liberals, however, is that Ted Hsu is now the favourite in Kingston and the Islands, ahead of the Conservative challenger.

Elsewhere, the New Democrats are now projected to be ahead of the Tories in Elmwood - Transcona, NDP MP Jim Maloway's riding. The Conservatives had been projected to win the riding on Friday.


  1. For the purposes of my own amusement, I'm going to try and predict the direction of Nanos changes tomorrow:

    CPC - Down
    LPC - Up
    NDP - Up
    Green - Up
    Bloc - Up

    (This is based on my current guesswork as the to daily polling numbers and has a really good CPC day falling off)

  2. Hi Éric-

    Thanks so much for the site and the work you've been putting into it.

    A question about today's projection: your riding projection for Sault St. Marie has the NDP at 40.2 and the Tories at 39.5, yet the seat is highlighted for the Tories...?

  3. Ugh, a mistake. Thanks Anon.

  4. Hello Eric,

    Great site you have and thanks for all your hard work. I noticed the following article which may imply the Quebec riding of Abitibi-Baie-James-Nunavik-Eeyou is up for grabs.

  5. I've given Saganash the "star candidate" bonus, but he has a long way to go.

  6. Hello Éric,

    Which Léger poll are you referring to in this post? I can't seem to find it anywhere.

    Christine D.

  7. does the Nanos rolling poll with the Bloc at 36 and CPC at 25 have any credibility at all?

    How strong is the echo effect?

    The Leger report has the green polling higher than their 2008 election results. To me it kinda of invalidates the overall accuracy of the poll as 2008 the environment was a much bigger issue and was a perfect storm for the Green vote.

  8. Christine,

    It was reported on in the Sun Media chain. I'll have links and full details up soon.

  9. If the Nanos poll holds at Bloc 36 CPC 25 . (Along with the huge CROP Quebec poll having Liberals at 10%) Mr. Harper could offer to debate Duceppe one-on-one in French instead of the current 4 way melee.

    It wouldn't be accepted but would marginalize the Liberals in Quebec much the same way Ignatieff is marginalizing the NDP in the ROC by pushing for a one-on-one with Harper.

  10. Green vote is way to high in the Leger poll, not going to happen. it will be down to 4 or 5% again. They have no profile this election.

  11. Any poll that gives the CPC gains in Quebec, with BQ support below 45 is suspect. I only say 45, as some of the ridings can eventually switch, but not enough to drop them below that point without some seriously damaging scandals coming out against the BQ.

    Larry Smith all but harpooned the CPC's chances for gains in Montreal, and possibly overall in Quebec, due to his poor choice in advisers.


  12. Is Nanos the only firm doing these daily tracking polls? It seems surprising that EKOS, Angus-Reid and Harris Decima seem willing to take a back seat to Nanos during what is arguably, their best opportunity (election time) to advertise their brand. The overall silence from the other pollsters is very odd.

  13. As a Green it is a frustrating election. Nanos, iirc, is asking if you support the usual 4 but requires you to state you will vote Green thus biasing the numbers down. The media seems to only report on the Greens over the debate issue, ignoring daily messages listing our proposals for high speed rail, reducing job taxes, etc. I wonder how different the election would be if the Conservative daily talking points were ignored, and pollsters didn't list them.

    Sigh. Could be worse I suppose.

    Meanwhile it is interesting to watch how the numbers shift here. Does the fact Nanos is polling daily cause them to have a stronger pull on the overall figures, or does the weighting reduce them (or increase the others) a bit?

  14. Nanos being the most active, they are taking a larger place in the projection than they normally would. Should be rectified as the other pollsters weigh in.

    But I'm only including Nanos numbers that don't overlap with other daily polls.

  15. Leger has Green as one of the visible choices in their online question. Nanos asks over the phone without prompting ANY names. I think that when the dust settles - the Greens will get a percentage more in line with the unprompted numbers (i.e. 4%) but that difference won't be people who vote for other parties - they just won't vote at all. So if you took the Leger poll and adjusted it for the Green/Other overestimate - You would probably get Tories at 38%, Libs at 27%, NDP at 19% and Green other at 5%.

  16. So far the Liberals haven't imploded yet.

    No doubt the incredibly positive media coverage (and negative coverage of our PM) has been helpful to their cause.

    I also notice that Jack Layton is getting pretty mediocre coverage as well.

    Still the official media story lines just don't match the polling data.

    Harper has gone up. Ignatieff has been flat.

    Once we start getting some IR and AR numbers showing big CPC leads a sense of panic will set in amongst the Liberals.

    The infighting well return. The echo effect in Quebec will see their numbers collapse and shift to the Tories. Morale will decrease.

    A great debate performance by Ignatieff could keep them afloat for a week or so.

    But by the end of the campaign we'll probably see numbers like - CPC 43, LPC 23, NDP 21, BQ 9, Green 4.

  17. Its been what? A week? No major polls from A-R, I-R, EKOS, Harris? Weird.

    Are they even planning on using tracking or are they just planning on big sample dumps?

  18. Kevin

    EKOS comes out every week ... used to be Thursday now is Friday.

    Somehow it is the Toronto Star using pro CPC pollster AR.

    If there was Angus Reid poll done that showed CPC 46 Liberal 20 I don't think that the Star would be keen to publish it until Ignatieff has a chance to get at least another week of campaigning. The media would start talking about the Liberal leadership campaign.

    The Liberals seemed to orchestrate a fund raising push as the government was defeated. I would do the same if I was in the Liberal back room to give the impression of momentum.

    The fund raising stops (or at leasts will slow to a trickle) if a Liberal election disaster seems to be unfolding.

    The $1,000 Donors will hold back until they see if there will be a NDP-Liberal merger of equals into the Social Democrat party.

  19. Thanks for running

    There's a new method of measuring elections using markets, and you might wish to take it into account:

    Sounds quite interesting and more likely to be accurate, because people are actually
    putting their money where their mouths are.


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.

Note: Only a member of this blog may post a comment.