Friday, April 22, 2011

Tight race for second as Conservatives coast untroubled

The four newest polls all show the same thing: the Liberals and New Democrats are now statistically tied at the national level while the Conservatives enjoy an unassailable double-digit lead.
These polls from Angus-Reid, EKOS, Forum, Ipsos-Reid, and Nanos add more fuel to the fire as to whether this will be a historic election for the NDP and a historic defeat for the Liberals. But while all of this goes on at the margins, the Conservatives remain on track to win another minority - or even a majority.

The Ipsos-Reid poll for Postmedia grabbed the most attention yesterday, as it put the Liberals at an incredibly low 21%, behind the NDP at 24% and the Conservatives at 43%. There is little in other polls to argue that the Conservatives really are that far ahead, and when you compare it to Ipsos's last poll taken April 5-7 you see that the variation has been within the MOE.

That isn't the case for the Liberals and NDP, however, who are down and up five points each.

Far more interesting are the EKOS and Nanos polls, and how they have shifted over the last few days. EKOS's last poll was taken April 15-17, and since then the NDP has gained 4.7 points, mostly at the expense of the Conservatives (down three). In Quebec, the NDP is up six points while the Bloc is down 4.7.

Nanos shows something similar, with the NDP up 6.4 points over the last three days at the national level. However, Nanos has support being drawn from both the Tories (down two) and the Liberals (down 4.1). In Quebec, Nanos has the Bloc down 4.4 points and the NDP up 3.3. But unlike EKOS, Nanos still has the Bloc ahead of the NDP in Quebec.

All of these polls also show that the New Democrats are doing very well in British Columbia, though primarily at the expense of the Liberals. So let's take a look at how polling has evolved in the West Coast province since the start of the campaign.

Generally speaking, the campaign in British Columbia has been relatively flat. Or at least it was until the 16th and 17th, the weekend immediately after the debate that coincided with the BC NDP's choice of a new leader.

The Conservatives have been humming along at a good pace around 42%. For much of the campaign the Liberals and NDP have been locked in a battle for second, but the Liberals had the upper-hand.

But over the last weekend that shifted, and the Liberals dropped from roughly 25% to around 19%, generally where they were in 2008. The NDP has jumped up to over 27% support in the province, meaning the NDP is set to improve upon its 2008 performance.

You can click on the chart to the left for the detailed breakdown of all polls conducted in British Columbia since the campaign began.

Along with the Liberals, the Greens have been struggling in what is supposed to be their best province. From the 12% range where they were when the campaign began, the party has sunk consistently. In today's Nanos they were at only 4%, but over the last few days they've averaged about 7%. That is still worse than 2008, and does not bode well for Elizabeth May's chances in Saanich - Gulf Islands. Gone are the days when the Greens routinely polled over 15% in the province.

While the polls haven't changed too much in British Columbia since the 2008 election, some of the personnel changes that have taken place in the province (Keith Martin, Bill Siksay, Jay Hill, Chuck Strahl, Stockwell Day, etc.) mean the people that British Columbians send to Ottawa won't be the same. And the retirement of these MPs, some of them in some close races (Esquimalt - Juan de Fuca and Burnaby - Douglas in particular) does make the potential outcome in the province difficult to predict.


  1. Not sure if Ipsos should be taken at face value. They tend to inflate CPC numbers....

  2. Eric,

    Can you explain why the Ipsos numbers are so off the tracks compared with other polls? Since the beginning of the election Ipsos have the Conservatives always 40% and higher...since they publish polls for the National Post and Global, are they calling high conservatives riding to get their number??

    2nd question: What is the difference between your projection and these projections?


  3. I think I have a serious issue with your contention that the NDP would actually lose seats in BC. The party is poised to gain from a lackluster federal Liberal campaign, and a re-alignment of progressive votes. I believe there are several seats in play for the NDP...Vancouver Island North; Surrey North, Newton North Delta, Esquimalt Juan de Fuca...collapse of Liberal and Green votes in BC should put these 4 within a hair of the NDP column; and all of them are not particularly friendly to Tories.

  4. Further to my comments on BC. In 2008, the NDP got 28% of the vote; they're poised for more now. Most polls have them above that, within striking distance of the tories who are mired in the low 40's. But they had 46% in 2008 - this means they're poised to drop a couple seats too. I live here in BC, and there's a lot that a poll can't report.

  5. On your site you used to have a break down of all of the pollsters and the amount that they tended to lean towards one party or another. Is that still there?

  6. a notable trend you haven't mentioned: there's been a surge in the number of undecided voters.

    is it possible that what's happening isn't so much of a surge in support for the NDP, so much as softening support for everyone else? with the other parties starting to have voters question them... a consistent NDP would show up as a rising NDP?

    or am I reading too much into the undecideds?

  7. Ipsos-Reid is "coin of the realm" in terms of consistent quality, though Nanos has also made a name for itself (though its last 2008 poll was out, except for the very last day of polling). The three Ipsos polls taken from the beginning of the campaign to now have been really consistent for the Tories 44%-41%-43%.

    On BC, I would have thought that the Conservatives shoudl be close to winning a few seats also. Esquimalt-Saanich, Vancouver South...maybe even Quadra can all go Conservatives though I will admit that the tight NDP seats might stay with the party.

  8. Eric, would you be able to provide a list of the ridings most likely to be gained by the Conservatives due to the Liberal/NDP vote split? My intuition would be that the net result of the Liberals and NDP polling at the same levels would be a massive Conservative majority due to vote splitting, particularly if the NDP starts surging in Ontario.

  9. Ipsos had the CPC at 37% on Oct 2, 2008 ... 12 days before the last election.

    They then had them at 34 to the Liberals 29 on Oct 9.

    The CPC actual did better than IR polled them last election.

    The 43 might be a bit under

    AR and Leger with 16 and 14 point spreads is definitely bringing the IR 19 point margin back into the realistic range.

    The Mansbridge - Harper interview in 2008 where Harper spoke the truth that a CPC majority would be held in check by the Liberal public service and judiciary is very similar to the Mansbridge Ignatieff 2011 interview.

    Both leaders were honest. In 2008 cost harper a majority in 2011 Ignatieff will give Harper a majority.

  10. Any information on interior BC ridings? Kootenay - Columbia and Kamloops - North Thompson both have retired Conservative MP's no longer running. I'd keep my eye on both if the NDP continues doing very well in BC. Both of these areas are quite strong for the BC NDP provincially, and became massive Conservative victories mainly out of the personal popularity of the Tory candidates and the afterglow of the Reform party days when the NDP lost their strength in the 90's. Both seats are definitely leaning conservative, but if the NDP continues doing well... these ridings COULD come close to getting into play.

    When the NDP won BC Southern Interior after the Tories lost their candidate, the NDP established a firm hold on the riding afterwards. It could happen to either of these two interior ridings... but I admit it's still a longshot.

    That said, I think your numbers for Kootenay - Columbia are higher than they really are... Wilks is more at the 45-49 percent range IMO. Apparently has been missing local community debates, the NDP candidate is a former Mayor of Invermere, and Jack Layton has paid two visits to the riding already. The NDP is definitely focused on it this time... but the Greens and Brent Bush would likely split the coalition of possible voters in the end of the day.

  11. Ya, and EKOS seems to deflate them. It seems like Nanos is the most stable (and most reliable?).

  12. Steve,

    That site appears to do projections for each individual poll. I don't know anything about their methodology. My projections are based on an aggregation of all polls.


    That is still somewhere in the archives, but it hasn't been updated for a very long time and so isn't current. It was also a very rough calculation, and I decided it wasn't accurate enough to maintain.

  13. As someone who correctly predicted the NDP surge (indeed that they would pass the Liberals) ages ago i'm going to comment on the BC situation.

    CPC will pick up Vancouver South, Vancouver Quadra, Newton-North Delta, and Esquimalt Juan De Fuca from the Liberals.

    Only Hedy Fry will survive 2011.

    Burnaby-New West, Burnaby-Douglas, BC Southern Interior, Surrey North, and Vancouver Island North will all be razor close CPC-NDP contests with the edge on the incumbents.

  14. "AR and Leger with 16 and 14 point spreads is definitely bringing the IR 19 point margin back into the realistic range."

    What are you talking about?

  15. Eric - in case you did not receive the tweet I sent you - a new riding poll conducted in Brome-Missisquoi.

    Bloc 32 (-3.2 from 2008)
    NDP 26 (+17)
    Liberals 26 (-6.8)
    Conservatives 11 (-7.7)

    Looks here that the NDP is mostly cleaving support away from the other federalist parties - and not from the Bloc.

    of note, according to Segma polling chairman:

    «Le fait saillant de ce sondage est que le candidat NPD (Pierre Jacob) a vu ses appuis passer de 19% la première journée (le lundi 18 avril) à 33% lors de la quatrième journée du sondage (le jeudi 21 avril), ce qui le plaçait tout juste devant la bloquiste Christelle Bogosta pour cette seule journée. Pendant la même période, la bloquiste a vu ses appuis fondre de 19%!», souligne le sondeur Raynald Harvey dans son rapport. «C'est presque incroyable, ajoute ce dernier en entrevue téléphonique. Ça peut jouer de jour en jour, mais comme ça, c'est exceptionnel!»

  16. Anonymous said...
    Any information on interior BC ridings? Kootenay - Columbia and Kamloops - North Thompson both have retired Conservative MP's no longer running.

    Cathy McLeod is the Conservative MP for Kamloops, first elected in 2008, and she is running for re-election. The BC ridings where the incumbent Conservative has retired are: Prince George-Peace River, Kootenay-Columbia, Okanagan-Coquihalla and Chilliwack-Fraser Canyon.

  17. Re: Difference between Ipsos Reed, Compas, etc showing a majority gaining lead for the Tories while others show a ten(ish) percent lead.....

    By way of disclaimer roots are in US election polling, so these comments may not ally.... But here goes....

    In the last Canadian election only 59% of Canadians voted, so in doing a poll it certainly makes sense to try to filter your sample down to the portion of the population that will actually vote..... The classic "likely voter" poll in US elections

    Turnout used to be a lot higher in Canada and until recently the age, income, etc of the voter did not make a huge difference on voter turnout patterns. Recently however turnout patterns have been somewhat shifting so the older voters are more likely to vote, younger votes less so.... On on down the line based on income, education, etc.

    Canadian polling does not quite seem to have settled out it's methodology on the "likely voter" issue.

    For example, NANOS stratifies for age, so he mathematically adjusts his results to have, for example, young people aged 18 to 25 reflect in his numbers representation equal to their actual share of the population. This is perfectly valid if one is doing a poll of adult Canadians but may not be a good prediction of actual voters on election day...

    Older Canadians actually vote in numbers in excess of their proportion of the population, so a "likely voter" would overrepresent older voters and underrepresented younger ones...

    In 2008 the average of all polls understated the Tory lead by about 3%, mostly (I think) do you making the samples representative of the whole population rather than the actual likely electorate

    Screening down to the firmly committed voters like jCompas and IR did "could" be acting as a type of likely voter screen, thus reflecting a better outcome mfor the Tories.

    I might be wrong, but this is what it looks like kkto me from an admittedly US polling centric perspective

    The Vorlon

  18. Would be hard to see the Liberals more than 1-2 seat in BC, after all they were at 19% last time and still won 5, and are up a bit this time around. I dont know if that means theyll gain any but theyll just hold 4 of what they have now most likely. Havent seen too many polls that have Conservatives over the 45% they got in BC last time and certainly none over 50.

  19. Eric, where's my crack?! I'm jonesing hard here :-)

  20. There is one thing going on that I find rather odd??

    Apparently the advance polls are seeing action at unprecedented levels. Reports of waits for an hour or more are common.

    Now how does that tie in with the pollsters who seem to be predicting a pretty quiet vote?? Certainly they don't seem to report any great excitement ??

  21. Peter i'm glad you've noticed all the voting!

    In 2008 the ridings with the highest advanced voting were the ones in which the CPC did the best.

    Having the best GOTV effort of all the parties this is a very good sign for the CPC.

    Did you know that you could actually vote for quite some time now ?

    You just mail away for a special ballot!

    On the CPC's website they have all the voting information and even a voting hotline to help lock in the votes of their supporters as early as possible.

    Don't be surprised if the polls are underestimating the Tories like they did in 2008.

    Its hard to capture the effects of a GOTV machine.

    In '04 and '06 the Liberals big red machine was underestimated. IN '08 it was the Tories.

    Things have just gotten even better, while the Liberals have gotten even worse.

    Don't be surprised if the Liberals are wiped out and the Tories surprise you on e-day.

  22. Vorlon, true but young voters in Canada 18-34 are more likely to vote Conservative at 37% over other parties.

  23. "Apparently the advance polls are seeing action at unprecedented levels. Reports of waits for an hour or more are common.

    Now how does that tie in with the pollsters who seem to be predicting a pretty quiet vote?? Certainly they don't seem to report any great excitement ??"

    Apart from the obvious... Friday being a stat holiday meaning (a) people have time to vote and (b) people have nothing else to do?

  24. You can get a better guess of current national levels just from the new Ekos poll. They had the libs and ndp right on in 2008 but the Tories a bit low. In that election the Tories surged in the final weekend so the Ekos numbers were right on. Maybe a point low on the Tories so adding one point on the current national number we get Cons 35 NDP 28 Libs 24
    308 is adding the Cs and Libs to be higher than they 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.