Tuesday, September 15, 2009

New HD Poll: 4% Conservative Lead

Harris-Decima has released a new poll, and thanks to the Toronto Star for publishing the results in full.

The poll was taken between September 3 and September 13 and involved just over 2,000 respondents. This is part of a 'rolling poll', as the previous HD poll was taken between August 27 and September 6. For that reason, this poll will be reduced in weight by 36% since part of it has already been represented in the model.

The national results:

Conservatives - 34%
Liberals - 30%
New Democrats - 15%
Greens - 10%
Bloc Quebecois 9%

Once again, Harris-Decima provides us with a poll to counter a suspicious Ipsos-Reid poll. I wonder what kind of inter-pollster rivalries exist.

In British Columbia, the Tories have a solid lead with 36%. The Liberals and NDP are tied at 23% and the Greens have a strong 17%. Alberta is as you'd expect, but in the Prairies the Tories are at 45% and the NDP is second at 35%, with the Liberals in third at 18%. That is a very strong result for the NDP.

In Ontario, the Tories don't have the 10-point lead Ipsos-Reid gave them. Instead, the Liberals are ahead three points, 38% to 35%. The NDP is alright at 15%.

In Quebec, the Bloc is doing well with 39%. The Liberals are doing good too, with 30%. The Conservatives are at 16% but the NDP has fallen to 7%, just one point above the Greens. How much is Thomas Mulcair worth to the NDP?

In Atlantic Canada, the Liberals have dropped to 32% with the Tories at 30% and the NDP at 24%.

This poll would give the following seat totals:

Conservatives - 126
Liberals - 104
Bloc Quebecois - 51
New Democrats - 27

So, something similar to the 2006 result. Significantly, the Liberals and the NDP could out-vote the Tories.


  1. Re: inter-pollster rivalries. Well, an awful lot of today's big pollsters at one point worked for Angus Reid. And they all seem to disdain his methods.

    I met Michael Marzolini about a decade ago and he told me that they had a rule in-house: if an analyst ended up with the same results as Reid they were to throw them out because they were obviously wrong.

  2. Huh. And they were the closest pollsters in both the Canadian and Quebec elections last year.

  3. I wonder if Marzolini is referring to Ipsos-Reid or Angus Reid?

    In any case, I think that the key number after the election is whether the Liberal+NDP seat count exceeds the Tory seat count. If that happens, then I think that there is a very strong chance that the Liberals would form a minority government with some sort of accord with the NDP.

  4. Which, even if they won fewer seats, might be a better position for the NDP to be in.

  5. The full details and methodology for the HD poll can be found here.

  6. No, that is the poll that was taken last week. They're a little slow with the updating.

  7. BC Voice of Reason15 September, 2009 17:24

    Eric... I love your site and think you provide as unbiased presentaion of the poll results as exists.

    What is with "Once again, Harris-Decima provides us with a poll to counter a suspicious Ipsos-Reid poll." ?

    Ipso-Reid , according to your bias research favour the CPC by 3% and goes -.3 for the Liberals.

    the Harris decima boys are -2.5 for the CPC and -1 for the Libs.

    So this 4 % spread real is more like 5.5% and the 9 spread in the I-R poll is 5.8% so basically with your research these polls are telling us the exact same thing and are not out-liers.

    The rolling H-D poll contains 1/3 if the data taken prior to Ignatieff crossing the Rubicon to take on the Holy Harper Empire.

    There was no trend data provided with the H-D poll. I think that the H-D poll results are worse news for the Libs than the IR poll.

    The last H-d Poll you have in your stats tables had the Liberals with a 35-31 lead. Way back in June.

    I hope you are not starting to spin your results.

  8. Not at all. Just that no one has shown such a spread as IR, and the one before it was even larger. And on both occasions we received a poll on the same day with very different results.

  9. This was 1998 so Marzolini was referring to the old Angus Reid before he sold out to Ipsos.


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.