Friday, March 25, 2011

Big numbers, little real change in new Ipsos poll

A lot of ink was spilled over the newest Ipsos Reid poll released yesterday through Postmedia and Global Television, as it put the Conservatives ahead by, oh, 19 points. But, looking at it more closely, it isn't nearly as remarkable as it might appear at first glance.

Compared to Ipsos's last poll conducted in early March, the Conservatives have gained three points and now lead with 43%. The Liberals, meanwhile, are down three points to 24%.

Neither of these two shifts are statistically significant, though that doesn't necessarily mean they are false. But Ipsos had the Conservatives at 43% back in late February, and then dropping down to 40% in early March, so this isn't a breakthrough poll for the Conservatives or any sort of Ipsos-trend. Their numbers are just wobbling back and forth.

The New Democrats are unchanged at 16%, while the Bloc Québécois is down one to 10% and the Greens are up one to 6%.

Ipsos Reid also asked how likely people are to vote. About 54% said they were absolutely certain to vote, and I believe them. Another 19% said they were very likely to vote, but at least half were lying. And then you have 10% saying they were somewhat likely to vote, while 5% said they were not very likely to vote. Another 12% of honest people said they were not at all likely to vote.

Of those who are absolutely certain to vote, 45% will vote for the Conservatives. Only 23% will vote for the Liberals, 15% for the NDP, 10% for the Bloc, and 7% for the Greens. Of course, the sample size of those likely to vote is about half of this entire sample, increasing the sampling margin of error to over +/- 4%.

Regionally, there is little change in Ontario. The Tories are up, however, gaining three points to lead with 46%. The Liberals are down two to 30%, while the NDP is up one to 16%. This is an important number for the Conservatives.

In Quebec, the Bloc has dropped five poins but still leads with 41%. The Conservatives are up 10 points (a statistically significant gain) to 25%, and lead the Liberals who are down three to 18%. The NDP is down one to 13%.

In British Columbia, the Conservatives are steady at 50%. While a big number, we've seen other results with the Tories about this high in the province. The Liberals are down six to 22%, the NDP is up one to 20%, and the Greens are up four to 7%.

The Conservatives lead in Alberta with 54% (-9), followed by the Liberals at 23% (+12) and the NDP at 17% (-2). Those opposition results look a little high for Alberta.

But then the government's numbers in the Prairies are high as well: 66%. That's a gain of 15 points. The Liberals are down 17 points to 18% while the NDP is steady at 13%. Chalk that one up to the 54 people surveyed in the two provinces.

Finally, in Atlantic Canada the Tories have gained six points and lead with 41%. The Liberals are down four to 28% and the NDP is up four to 20%. This isn't the first time we've seen the Conservatives ahead out east.

Not surprisingly, this one poll is enough to give the Conservatives a majority. With this poll only, ThreeHundredEight projects 25 seats for the Conservatives in British Columbia, 26 in Alberta, 24 in the Prairies, 63 in Ontario, 11 in Quebec, and 17 in Atlantic Canada for a total of 167, a gain of 13 over Ipsos's last poll.

The Liberals win five seats in British Columbia, one in Alberta, two in the Prairies, 28 in Ontario, 12 in Quebec, and 11 in Atlantic Canada for a total of 60, a drop of 11.

The Bloc Québécois wins 51 seats in Quebec, six fewer than last time.

The NDP wins six seats in British Columbia, one in Alberta, two in the Prairies, 15 in Ontario, one in Quebec, and four in Atlantic Canada for a total of 30, a gain of four.

So, a remarkable result for the Conservatives. But it is just one poll. We'll see how things change in the next 100-150 polls due to be released in the next six weeks.

Ipsos did look at a few other issues, including how Canadians view the budget. About 19% give it a thumbs-up, about 16% give it a thumbs down, and 48% shrug. Those terms are how Ipsos has effectively characterized it.  Surprisingly, even 42% of Conservatives only have one thing to say about the budget: "meh".


  1. Those responses to the budget are, I think, good news for the Conservatives.

    Since it was the budget that ultimately led to the defeat of the government, it will be difficult for the opposition parties to explain why they did that, given that most Canadians genuinely don't care about it.

  2. You have the Liberals gaining a seat in Alberta with this poll? Is that a typo?

  3. Did you read the poll? Conservatives 54% in Alberta, Liberals 22%.

  4. McGuinty steers clear of Iggy:

  5. Ignatieff made the first glaring mistake in the campaign.

    He looked the Canadian people in the eye and wouldn't answer a question about the coalition and said that was as clear as he could be.

    He just lost the support of all the MSM.

    Next weeks pools will have the liberals below 20 and the CPC 45-50.

    It is just shocking that the Liberals would not have expected the question and had a better answer.... Even saying that they were fully open to forming a government with the BLOC would be better than the contempt for the Canadian people showed in this none answer.

  6. Ignatieff's mistake is of the same magnitude of Kim Campbell's not wanting to discuss policy in an election. It is bigger than Dion not being able to understand a question in English.

    He is saying vote for me and I will use your vote as I see fit.

    Wow I am shocked that he was so unprepared and so bad on his feet. The bus tour talking to a few hundred Liberal partisans at a time obviously was not adequate preparation,

    Day -1 of a 36 day campaign, it couldn't have started worse.

  7. "Next weeks pools will have the liberals below 20 and the CPC 45-50."

    I'll hold you to that. You're entire credibility hinges on this awkward declaration being true. Or at least, what was left of it...

  8. I expect the projected coalition of Liberals, BQ and NDP will be one of two things that the Conservatives will talk about. The other being the budget.

  9. Well certainly some of our pollsters are very wrong and some of them are very right. There seems to be a larger-than-MOE sized difference between Harris/EKOS and some of the others like IPSOS.

    I think the idea that budget will be THE issue because it just happened to be the item the government was defeated on, will prove to be incorrect. It's not as if the other parties elected to vote in support of the government on the other stuff like contempt so they could vote them down on this item instead. This just came up first. So they'll still campaign about the others things, (Rightfully so) and Harper will presumably respond to that.

  10. BC,

    The problem for Iggy is that there's no good answer. If he says yes, than his support may bleed all over the place to the Tories and the NDP. If he says no, then he basically writes off any chance of forming a government after the next election.

    But you're right, the Liberals are going to have to develop an answer to that question, because it's going to be asked a lot. And if they can't answer it, that will become the storyline for the opening weeks of the election.

  11. Eric any idea when EKOS is coming out?

    Thanks, Earl

  12. Ipsos is currently the only pollster putting the Conservatives in this range. a 16 point lead in Ontario while Harris-Decima has the parties in a dead-heat with a larger sample size? Come on. The Ipsos numbers are veering into the comical, and I would say the same thing if they had the Liberals up at 40%.

  13. "But you're right, the Liberals are going to have to develop an answer to that question, because it's going to be asked a lot. And if they can't answer it, that will become the storyline for the opening weeks of the election."

    Harper was just as willing to form a coalition with the "socialists" and "separatists" in 2004. Any minority government must effectively rule day-to-day in "coalition" mode. To demand that Ignatieff should have to answer this and Harper should not, is preposterous.

  14. I'll hold you to that. You're entire credibility hinges on this awkward declaration being true. Or at least, what was left of it...

    He was was totally wrong on the coalition forcing an election and then total correct on the first poll after the election was forced.

    So even if I am wrong I would be 1 out of 3.

    If Mr. Ignatieff gets to 33% the Liberal party would be ecstatic.

  15. IR showing Alberta and BC with essentially the same numbers is hilarious (and hard to swallow). Do we need to start talking about "Fortress Saskitoba"?

  16. BC VoR,

    I know I would pretty much parade in the streets.

  17. Why is it always Saskitoba? I prefer Manatchewan.

  18. if this (ipsos) poll was the only one that had the conservatives so far in the lead, i wouldn't be cracking the champagne just yet, however there are many recent polls that have put the conservatives far ahead of their nearest rivals...leger,angus-reid,nanos research and this is the second ipsos poll in as many weeks to peg the conservatives at 43% so it appears that there IS something to see fact if there are any deviant polls lately they are the ones that have the conservatives at or below 35%....during this campaign look for some (deviant) polls that suggest a horse race...take this with a grain of salt - after all, the toronto municiple election was a "horse race" between smitherman and rob ford...(yeah!)

  19. Anonymous:
    "Ipsos is currently the only pollster putting the Conservatives in this range. a 16 point lead in Ontario while Harris-Decima has the parties in a dead-heat with a larger sample size? Come on. The Ipsos numbers are veering into the comical, and I would say the same thing if they had the Liberals up at 40%."

    Except none of that is true. The Ipsos poll shows a big lead, but then so do the latest Leger, Nanos and Angus Reid polls (and, if you insist on claiming that sample size makes polls more accurate - which is statistical nonsense - consider that the Angus Reid poll had a sample size in excess of 6400respondents).

    The Ekos and Decima polls are the odd men out here (although, I'm not sure you can say with a straight face that say that the Tories and Liberals are in a dead heat since they show the Tories being up by 7%, which puts the Tories in a better position than they were going into the last election). They may turn out to be right, but there's no reason, a priori to prefer them over the others.

  20. Carl,

    Last AR poll only had 1,021 respondents. Just fyi.

  21. Actually, BCVoR, those are just the last week. The March 10-16 numbers are in the release, but Ekos hasn't gone to any effort to draw attention to them.

    The numbers you quoted are March 17-24. March 10-16 looks like this:

    CPC - 34.1
    Lib - 25.7
    NDP - 16.4
    GRN - 10.4
    Bloc - 10.2

  22. Agreed. I won't allow comments like that again in the future.

  23. I'm sure we will have ample opportunity to comment on the Ekos poll in its own thread - but I have to say that once you get out of Quebec and Ontario - the numbers look absurd. Does anyone believe that Iggy-mania has swept Alberta bringing Liberal support there up to 32%??? I know some polls have had the NDP down a tad in Atlantic Canada but down to 6%? - come on.

  24. It was Tondra McCharles of the Toronto Star that was pressing Mr. Ignatieff on dodging the coalition question.

    She was trying to do him a a favour. She must have thought that he would actually have a good answer.

  25. Changed the way comments are entered. Better?

  26. Volkov,

    You're right, I was looking at the big one from late february. The point, though, remains valid.

  27. Joel-Denis Bellavance said on Powerplay that there will be a big CROP poll of 1,000 in La Press tomorrow that will be DEVASTATING for the Liberals with the NDP at 20% - and that 20% is "almost double" Liberal support - in other words federal Liberal support in Quebec is almost into single digits!

  28. There will be a Léger poll in Le Devoir tomorrow as well.

    Boy, this is going to be a busy six weeks.

    I will summarize all of the polls released today and during the weekend on Monday.


    That's the #1 election phenomena we need to be on the look out for.

    At what point during the election do the NDP and the Liberal numbers match up ?

    Then we can start asking whether Jack Layton intends to form a coalition with the Liberals to become PM.

    (Good God will that prospect ever bleed every single blue Liberal to the Tories.)

  30. I've heard some speculation that whoever wrote the story on this poll got some regional numbers reversed, so that the Conservatives are actually at 66% in Alberta and 54% in Sask/Man.

    Any chance this is correct? The numbers make a lot more sense that way...

  31. Unlikely, since I got the numbers from Ipsos Reid, not Postmedia.

  32. Eric, will you be working weekends at all during this campaign?

  33. Yes, but not at the same pace as during the week. For francophone readers, I will have articles in Le Devoir on Saturdays, and there will be an article on The Globe and Mail website on Sundays/Mondays. Those have to be written on Saturdays, so there is less time for me to update the site.

    But I will likely post one or two articles per weekend here on

  34. Unlikely, since I got the numbers from Ipsos Reid, not Postmedia.

    Okay, thanks for the info. What was the Alberta seat that the Liberals won in your projection?

    PS- Any change you could put up a graph of your seat projections over time, the same way you do for poll results?

  35. Two more big polls:

    Angus Reid says its CPC 39%, Libs 25%, NDP 19%
    Leger says CPC 39%, Libs 23%, NDP 19%

    Wasn't the Liberal "two election strategy" supposed to start with knocking out the NDP this time?? So far I think the Liberals have the worst of both worlds - they are shedding votes like crazy to the Tories and gaining ZIP from the NDP!

  36. Well, now we know why Jack Layton wanted an election. This may be the NDP's best chance to knock the Liberals out as the alternative to the Tories.

    If the Liberal strategy is premised on the claim that they're the only credible alternative to the Tories ("you have a blue door and you have a red door"), that strategy isn't going to survive the weekend.

  37. Isn't the missing 5% in the IPSOS vs EKOS poll for the conservatives, really the difference in Green party support? If the greens are at 6% the Tories are at majority territory. If the greens are at 11% the Conservatives are at 38%. Since Elizabeth May has been all but invisible for the past 2 years, I really think that 6% is more likely. (And I voted Green last time and live near where she is supposedly running for a seat)

  38. Shadow

    I prefer to call the Manley/McKenna/Tobin/Emerson/Paul Martin wing of the LPC as the Blue Dog Liberals.

    I know it is American but it is catchy and describes the the constituency. My Obama Democrat friends really are disgusted by the Blue Dog Democrats and their lack of commitment to the President's more socialist agenda.

    The Blue Dog Liberal leaders left the party and were replaced with Rae, Dryden, Dion and Kennedy.

    Manley was on CTV for budget analysis before the NDP came out against it and said it was a good budget.

    The voters that support the Blue Dog Liberals have to be bleeding directly to the CPC. The main complaint that they would have had was the social conservatism of the Reformers.

    After 5 years of government the tolerance and working with the visible minority rights and fairness in immigration has to make the Blue Dog Liberals why they have voted for the Liberals for the last 20-30 years.

    Recently the left has shown more intolerance and demands for conformity to their ideology than the right.

  39. Anyone disagree that the NDP beating the Liberals is now a more likely outcome than the Liberals beating the Tories ?

    Convergence! Criss Cross! Roadhouse!


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.