Friday, December 17, 2010

Liberals and Conservatives sink in new EKOS poll

In EKOS's last poll of 2010, the Conservatives and Liberals have both dropped to the benefit of all three of the smaller parties.Compared to EKOS's last poll a week ago, the Conservatives have dropped 1.7 points and lead with 32%, similar to what Harris-Decima reported earlier this week. The Liberals, however, are down 2.7 points and trail with 26.5%.

The New Democrats are up 2.7 points to 17.1%, while the Greens are up 0.5 points to 10.9%. The Bloc Québécois stands at 10.6% nationally.

Down two points from a week ago, 14% of respondents in this automated telephone poll were undecided. EKOS reported a week earlier than usual because of the holiday season - we won't hear from them again until after the New Year.

The Conservatives are holding steady in Ontario, and have a narrow lead over the Liberals with 34.6% to 32.9%. That's a drop of one point for Michael Ignatieff's party. The NDP is up one point to 17% while the Greens are steady at 12.4%. The Conservatives lead in Ottawa with 42% to the Liberals' 38%, but are trailing in Toronto at 37% to 39%.

In Quebec, the Bloc is up four points and leads with 41.5%. The Liberals are down five big points to only 18.1%, a very low number for them. The Conservatives are up one to 17.5%, while the NDP is also up one, to 11.5%. The Bloc leads in Montreal with 41%, followed by the Liberals at 23%.

The Conservatives are down six points in British Columbia but still lead with 35.9%. The NDP is up one to 22.8%, while the Liberals are down five to 20.3%. The Greens made the biggest gain, up seven points to 16.4%. The Conservatives lead in Vancouver with 38% to the NDP's 20%.

Undoubtedly due to the small sample size in Atlantic Canada, the Liberals have dropped 21 points to 36.9%. The Conservatives are up 14 to 33.7% while the NDP is up nine to 22%. These are all more along the lines of what we'd expect out East.

In Alberta, the Conservatives are down 17 points to 45.6%, followed by the Liberals at 25.9% (+8) and the NDP at 15.4% (+9). The Conservatives lead in Calgary with 53% to the Liberals' 28%.

Finally, in the Prairies the Conservatives are holding steady at 44%. The NDP is up six to 24.7% and the Liberals are up eight to 22.7%.

The Conservatives would win 22 seats in British Columbia, 25 seats in Alberta, 20 seats in the Prairies, 46 in Ontario, seven in Quebec, and nine in Atlantic Canada for a total of 130. That's one fewer than the projection for EKOS's last poll.

The Liberals would win seven seats in British Columbia, two in Alberta, four in the Prairies, 43 in Ontario, 12 in Quebec, and 19 in Atlantic Canada for a total of 89. That's 11 fewer than last week.

The Bloc would win 55 seats in Quebec, three more than in EKOS's last poll.

The New Democrats would win seven seats in British Columbia, one in Alberta, four in the Prairies, 17 in Ontario, one in Quebec, and four in Atlantic Canada for a total of 34, nine more than last time.

There are a lot of similarities between this poll and the one released by Harris-Decima earlier this week. For example, the high Green number in BC and the (relatively) closer race in Alberta. It certainly adds fuel to the fire in terms of the debate between whether the Conservative lead has grown or whether it has remained about the same as it has been for most of the year.

I suspect this will be the last poll we'll see in 2010, but I will be updating the site as much as possible during the holidays. I won't lack material as I still owe you all November round-ups for polling results and best/worst case scenarios.

The next year should be very busy, as we'll have four provincial elections (Saskatchewan, Manitoba, Ontario, and Newfoundland & Labrador) and perhaps a federal campaign as well. British Columbia could head to the polls after their leadership races are over, and there are rumours of an early election in Alberta and Prince Edward Island. It'll be a big year.


  1. Eric, Merry Christmas! Thank you for all the work you've put into this site in 2010. Congratulations on your work with the Globe and Mail and other publications!

    It has been a banner year for you.

    All the best in 2011 and beyond.


  2. I was wondering what programs you use to make your graphics! Thanks

  3. I use OpenOffice. Simple, free, and effective.

  4. Should be interesting to see if the CPC gets its usual bump after Iggy plays the war drums:

    My guess is the 40 -45% of the electorate is not that unhappy with Harper that they want, what they will see as an unnecessary election. I'm thinking if Iggy beats the drums loudly we get the CPC at or above 40%. Then what does Iggy do?

  5. When you read the interview in detail, it doesn't sound like Ignatieff was really beating the drums very loudly. The headline says more than he did.

  6. Eric you may be right and were I IGGY I wouldn't want to appear anxious for an election. Better to appear to be forced into one on principle that be baying for an election.

  7. Éric said...
    "When you read the interview in detail, it doesn't sound like Ignatieff was really beating the drums very loudly. The headline says more than he did."

    Agreed, it sounded like the sort of thing any credible opposition leader has to say. It's not like he can say "An election, are you freaking kidding me, have you seen my polling numbers?" (actually, that would be pretty funny). If it sounds strange coming from a Liberal leader, that just speaks volumes about their credibility heretofore.

  8. The Bloc's result in your article mentions 55 seats. There's most likely a typo on the image, as it says 56 at the moment.

    I love 308, thanks for your information. It's just this little fix, 309 just wouldn't be the same.

  9. Earl Ignatieff is basically saying what he was saying at the start of the year.

    No confidence in government, Canadians are tired of Harper and want an election, and we're ready to go if there is one.

    That's why the Liberals opposed this year's budget.

    (Except a handful of them stayed out of the house for every confidence vote and allowed it to pass.)

    Its just more posturing, more game playing.

  10. So Ipsos Reid and Angus Reid both have the Conservatives in the high thirties and rising, while Harris-Decima and EKOS both have them barely above thirty and falling. Strange. Let's see what the new year brings... Maybe the traditional New Year prorogation?

  11. "Liberals and Conservatives sink in new EKOS poll"
    This is great news for the Rhinoceros!

    I noticed that sensationalism too. I thought the MSM (besides QMI) was supposedly a bunch of biased leftists?

  12. Anon, thanks for noticing the mistake with the Bloc seats. I will fix it.

  13. problems with ekos poll

    part 1) sample is wrong

    First of all it is nice that Ekos shows the numbers that other pollsters hide. The sample size of the regions for example.

    One of the basis for statistical accuracy and margin of errors is getting a random representative sample. In providing the MoE EKOS pretends they have a random representative sample but Manitoba/SK/Alberta/BC has 6.9M eligible voters (in the 2008 election) and this sample has only 457 participants. Quebec has 5.6M voters and has 689 samples. The sample clearly does not represent the universe.

    Why? Is it time zones on the calls? Are they getting refused participation in the CPC strongholds?

  14. EKOS problems with survey Part 2

    They need some quality assurance and better systems people.

    41.5 % support for the Bloc in Quebec does not translate to 10.6% support nationally.

    Quebec has 24.0% of the countries voters. 41.5% support in Quebec gives them 9.96 % support nationally not 10.6% that they published.

    This would not impact Eric's seat projection model BUT it begs the question if EKOS can not do this simple extrapolation correctly how wrong are they on the rest of the adjustments that have to be make to "correct" their unbalanced sampling methods?

  15. I found this interesting, from an IR survey.

    42% say Harper deserves re-election, 58% says he does not.

    Its funny I often hear people say "well Harper only has the support of 1/3 of the country!! Over 70% of people are against him!"

    In actual fact the split is 42/58.

    Not a majority to be sure, but certainly a lot better than if often suggested.

  16. "41.5 % support for the Bloc in Quebec does not translate to 10.6% support nationally.

    Quebec has 24.0% of the countries voters. 41.5% support in Quebec gives them 9.96 % support nationally not 10.6% that they published. "

    This can easily be explained by the fact that the number of undecideds maybe lower in Quebec than in the rest of the country - so the BQ votes may appear bigger in the national sample.

    I'm not sure what your point is about the Ekos sample sizes in various regions - ALL polling companies will do some oversampling and undersampling to get enough interviews in each region. That's why polling companies do WEIGHTING - and I'm not going to get into a complex explanation of that - if you are curious I suggest you take a Stats 100 course at your local community college.

    If you actually saw the puny number of interviews per region in the Ipsos and ARS polls that you love to so much because they show better results for the Tories - you would be aghast.

  17. EKOS problems with survey Part 3

    It doesn't show the CPC in majority territory - and that's what BC and shadow want!


  18. Anon if your comment is nothing more than an attack on other board members, perhaps you should not post it ?

    Its a wonder it was approved to begin with.


    DL your explanation falls short.

    Undecideds would be irrelevant to national support, as they are factored out of each sub sample and then each sub sample is equally weighted to arrive at the national numbers.

    A far more likely explanation is that it was a transcription error, which isn't uncommon for EKOS.

    Or maybe they factor in differences in turnout for each provinces, which would increase the weight of the Quebec sub sample.

    (Quebec had higher than average turnout in 2008.)

  19. Please, no more discussions about EKOS's breakdowns and weightings. We're presented with the data, more data than anyone else gives us, but we don't know their formulas.

  20. And EKOS isn't alone - both the recent Angus-Reid and Nanos polls had national numbers for the Bloc that didn't fit with BCVoR's 24% Quebec weighting.

  21. Shadow - my posts are not personal attacks, they are criticism of your continual partisan twisting and spinning of the poll results that Eric kindly releases and analyzes on a semi-daily basis. I have yet to see any posts from you whatsoever, regardless of the pollster being discussed, that do not merely attempt to erode the NDP or liberals or cheer on the CPC.

    "Personal attack"? Grow a thicker skin.

  22. Anon

    Well said and absolutely correct.

  23. Anon its generally considerered trolling to make it your mission to criticize other board members.

    Direct your posts to the topic at hand and away from attacking others.

    I have yet to see you make a comment on ANY of the actual topics on any of these threads.

    As for this EKOS poll, I see nothing terribly wrong with it. Certainly not as bad as the one several weeks ago.

    Its EKOS, it is what it is. We all know what that means.

  24. "Its EKOS, it is what it is. We all know what that means."

    And thanks for proving my point.

  25. Proving what point Anon ?

    People from all parties have their gripes with EKOS.

    We have a dipper commentator on here who regularily rips into them over their unrealistically high green numbers.

    Not everything is about partisan politics here Anon.

    C'mon, drop the personal attacks and join in the conversation. Lots of off topic discussion too that drives our kind host crazy.

    Anyways, I think everyone is going to take a little break from politics for the holidays.

    You should too! This is no time to be so critical of others.

    Merry Christmas!


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.