Thursday, April 8, 2010

New Ekos Poll: 6.3-pt Conservative Lead

EKOS's weekly poll is out, and my apologies for the late post. As the Easter Weekend interrupted EKOS's polling, the sample size is much smaller than usual.Now, before you look at these numbers and say "huge MOE!", remember that most other pollsters survey about 1,000 people. Only EKOS, and recently Harris-Decima, go for 2,000+ people. So, though these numbers definitely look wonky, they shouldn't be any more wonky than results we see from Angus-Reid, Ipsos-Reid, or Nanos.

However, with such odd results from a pollster that is always consistent, one must come to the conclusion that EKOS's polling is consistent because of the sample size and not their method. For them to show such weird numbers in British Columbia and the Prairies leads me to believe that their method is actually a little worse than other pollsters. What they bring to the table is large sample sizes, which can overcome most methodological deficiencies.

Anyway, to the poll.

The Conservatives are up 1.4 points to 33.6%, a good result for them. The Liberals are up 0.3 points to 27.3%, while the New Democrats are down 0.1 points to 15.9%. The Greens are down one to 11.7%, and "Other" is down 1.3 points to 1.8%.

In Ontario, the Conservatives are up five points to 39.5%, while the Liberals are down one to 31.8%. This is a huge spread, and very worrisome. But then again, Harris-Decima had things the other way around yesterday. The NDP is up one to 16%.

In Quebec, the Bloc Québécois is up three to 39.4% and the Liberals are up two to 23.6%. The Conservatives are steady at 17% and the NDP is down two to 9.6%. Good results for the Bloc, bad for everyone else.

In British Columbia, well, the NDP is up five to 29.5% and move into first. Oddly enough, the Conservatives are unchanged at 28.8%. The Liberals drop 10 to 19.6%, and the Greens are up five to 18.2%.

In the smaller regions, the Liberals lead in Atlantic Canada with 38.1% (up five), while the Conservatives are up eight to 36.1% and the NDP is down five to 20.3%. In Alberta, things are relatively unchanged with the Tories at 55.7%, but in the Prairies we get a throw-away result. The Liberals gain 14 points and are at 35.2%, while the Conservatives drop 16 points to 32.8%. The NDP is up five to 18.6%. What is going on on the plains?!?

The Conservatives win 59 seats in the West, 57 in Ontario (can you imagine the Conservative Ontario caucus being as large as the Western caucus?), 7 in Quebec, and 10 in Atlantic Canada for a total of 133.

The Liberals win 18 in the West (seven of them in the miracle Prairies), 36 in Ontario, 15 in Quebec, and 20 in Atlantic Canada for a total of 89.

The Bloc wins 53 seats in Quebec.

The NDP wins 17 seats in the West, 13 in Ontario, and 2 in Atlantic Canada for a total of 32.

The Greens win 1 seat in the West (British Columbia).

At the very least, it's fun to see something different in an EKOS poll.

45 comments:

  1. With only 797 decided voters surveyed, this is clearly not the most reliable poll in recent EKOS history. However, it's amusing to note that the national Green figure of 11.7% is still more believable than last week's 12.7% based on over twice the sample size. Even that lower value could be high by as much as 1%. That said, Éric's predicted first Green seat is increasingly plausible.

    My favourite numbers are (as usual), the Saskitoba under-25s. None will vote Tory under any circumstances. Ummmm, with a 43.8% MOE since only five voters were polled. Yeahright.

    Looking at the shapes of the curves since the last election,
    * The Bloc is remarkably flat. They get 9.5% of the vote plus or minus dust.
    * The Greens show a very gradual but consistent upward slope. (Extrapolating, EKOS will report 100% Green preference in 2070. The commenters on sixhundredthirtytwo.com will predict that only half the voters will actually mark their ballots Green.)
    * The Dippers display a bit more variation but have flatlined at their post-election polling levels. A member of the party might see an upward trend but it would require a large magnifying glass.
    * The gentle Green climb is noise to the Tories and Grits. They trade substantial voter share with each other in lock step, but neither can climb to majority territory. They see a zero-sum game, and like the generals of 1916, they don't have the imagination to break the impasse. All they can see is a war of attrition, trying to convince the voters that the other party is even worse.

    Obligatory predictability for Éric: fall election. Small Tory plurality. First Green or Greens elected. Non-confidence vote. Grit government. No coalition, but Dipper/Green/Bloc support. Stability for several years.

    Advice to Grits, Dippers and Bloc: start planning for that outcome. E-day + 1 is no time to be deciding what to do next.

    ReplyDelete
  2. I was surprised to see that they actually published these results and just didn't wait till next week and do an even bigger poll then usual.

    Although the regional rsults are screwed up the overall national results are on par.

    ReplyDelete
  3. Man/Sask is really away off. No way can the poll go from Cons=48% and Libs=23-24% one week to Libs being ahead the next week-that's just nuts.

    ReplyDelete
  4. Topline numbers look like the usual EKOS results we'd see.

    Forget the provincial numbers.

    ReplyDelete
  5. If you compare the Ekos with the HD that's below it significant differences appear.

    I'd suggest this Ekos is very suspect. Well off the mainstream of other stuff lately. Eric may be right about their methodology depending on large polling numbers for accuracy?

    ReplyDelete
  6. Both Ekos and HD have the NDP essentially tied with the Tories in BC - this in and of itself means DEATH to any Tory majority hopes and would mean at least 3 or 4 NDP gains in BC - if not more.

    ReplyDelete
  7. Eric

    As reported on the CBC site Ekos did quite a bit of polling on the Afghan war. Interesting numbers, any hope you could create sort of a mini report for that?

    ReplyDelete
  8. I believe this is the 20th poll.

    The Liberal, numbers in Saskatoba, are not right.

    I believe the Liberal numbers in Ontario are too low, and the tories to high.

    In my mind a rogue poll.

    ReplyDelete
  9. Rogue poll ?

    Sure. Its EKOS. So the 6.3 CPC lead should probably be about three points higher. And that green vote will never materialize.

    As for the regionals just ignore them.

    Doesn't invalidate the top line numbers or the fact that Iggy is in Dion territory.

    ReplyDelete
  10. When people talk about the "20th poll" they are referring to the overall national numbers being way off...and for all the bizarre regional stuff - at the national level, this Ekos poll seems pretty credible (apart from the usual absurd over-estimate of the Green party). If you actuially look at the regional numbers on any national polls of about 1,000 - you REGULARLY see wild discrepancies. In Saskitoba - we have seen the NDP go up and down by 15% from week to week and BC has been a wild rollercoaster ride as well.

    Personally, I think its irresponsible to release any numbers based on a sub-sample size of less than 100.

    ReplyDelete
  11. "Doesn't invalidate the top line numbers or the fact that Iggy is in Dion territory."

    It also doesn't invalidate that the Tories are well below what they got in 2006 let alone 2008 and unless they get a majority in the next election - Harper is politically DEAD. There is no scenario whereby the Conservatives will be allowed to form another minority government. Its either majority or over to the opposition benches. End of story.

    ReplyDelete
  12. Éric: However, with such odd results from a pollster that is always consistent, one must come to the conclusion that EKOS's polling is consistent because of the sample size and not their method. For them to show such weird numbers in British Columbia and the Prairies leads me to believe that their method is actually a little worse than other pollsters. What they bring to the table is large sample sizes, which can overcome most methodological deficiencies.

    With respect, it doesn't work that way. An error in the EKOS methodology would appear as persistently skewed numbers, not variance in small polls.

    This sample is slightly smaller than the other pollsters' and the national numbers look credible. Beyond that, if it looks too strange to be true, it probably is. The survey was worth publishing for the uninterrupted data stream, but I'm sure Éric will weight it appropriately in his projections, especially at the regional level.

    ReplyDelete
  13. DL

    Have to agree, Three stikes and you're out !!

    ReplyDelete
  14. To note, Harper could pull a Diefenbaker and hold on with all his might. I don't know how strong the party organization is against the leadership, mind you, but I don't think its too much opposition. Harper's got a little dictatorship over there.

    ReplyDelete
  15. Volkov i'd imagine Harper would indeed want to stick around if he was replaced with a coalition.

    He's actually proven to be a very effective opposition leader. I'd assume he'd go around stoking public anger at the unelected coalition.

    His strident tone would only further endear him to the base.

    Media and opposition figures would be howling for him to resign as an MP and be drummed out of politics.

    That's up to him and what I imagine would be a leadership review vote.

    ReplyDelete
  16. DL: [T]his Ekos poll seems pretty credible (apart from the usual absurd over-estimate of the Green party).

    Shadow: And that green vote will never materialize.

    I do believe that I detect some anxiety. That's reasonable, under the circumstances.

    ReplyDelete
  17. We had four federal by-elections in November in ridings that should have been prime territory for the Green party - they got between 2% and 4% in each of them. That speaks VOLUMES

    ReplyDelete
  18. Shadow,

    If the Libs/NDP, have more seats combined than the conservatives, how much anger do you think there will be if they come to a working arrangement together?

    Will that not represent the will of more Canadians.

    Harper should not be too strident, because that does not serve him well.

    The base may love Harper, but that is what you have now, and can not seem to expand on it.

    I dont think IMHO lightning will strike twice.

    I agree with DL, if Harper can not garner a majority, he is done.

    ReplyDelete
  19. Shadow,

    He could hound an "unelected coalition," unless of course the coalition wasn't unelected at all.

    If he did lose to a coalition, unelected or not, he'd be up for a leadership review. At that point, its down to whether or not the party wants more of Harper.

    ReplyDelete
  20. Volkov its up to Harper.

    He will always get the required 50% +1 vote to stay on.

    Its about what threshold he sets for himself.

    For anyone to say he is "done" w/o a majority is wishful thinking.

    Plus if this is a situation where NDP + Liberal = >CPC but <Majority the BQ could double cross Iggy and he could last about as long as Joe Clark did.

    ReplyDelete
  21. Shadow,

    it is always funny reading how Harper is "done" and may favourite running scared.

    Aside from the Party Financials.

    Aside the win of 2/4 50% in 2009 (in Quebec too!)

    Aside a gap of 3-6 PTS the Liberal leader may keep his job if he does better than the worst results in 150 years.

    This is too funny.

    Last 3 months EKOS has never polled CPC below 30%. Liberal above 30 2x.
    Fear the Liberals....lol
    (http://canadanewsdesk.com/polls/?p=Ekos&t=month3&l=&e=&n=All)

    ReplyDelete
  22. I'll say it again folks. Canadians will be given a clear choice between a coalition and the CPC. Harper and the CPC will make certain of it. While Harper may not get his majority I doubt that he will do worse then the present. So tell me Peter, and Volkov and others, how can the second and third parties legitimately claim to govern if they combined have fewer seats that the CPC. If my assumption is wrong then I concede that the way might be open for Iggy and Layton to presume to try and usurp power despite Iggy's election protestations to the contrary.

    The anti-coalition campaign may be a winner for Harper or a loser but I'm convinced he will play that to the hilt. So let's play what if and not sling around accusations. If Harper wins 140 to 150 seats but no majority, what happens. If he wins more seats than the LPC but less than the LPC and NDP do we get some working agreement either tactic or formal knowing they need to Bloc to survive as would Harper. This is the stuff interesting speculation is made of. BTW I'm not discounting the LPC gets more seats than the CPC, I'm just assuming that would mean a Liberal government.

    ReplyDelete
  23. Shadow,

    I don't think Harper would stay on if he got less than 70%, IMHO. If it was just 50+1, that's a sign of a split in opinion over him, and a drastic lapse in confidence. He'd resign.

    ReplyDelete
  24. "So tell me Peter, and Volkov and others, how can the second and third parties legitimately claim to govern if they combined have fewer seats that the CPC."

    Why would I tell you? I've been pretty clear and consistent in saying that the only way such a coalition would be legitimate is if the combined seat total was more than the Conservative seat total, and at least by 10 seats.

    ReplyDelete
  25. Thanks Volkov. Just interested in starting a discussion. Sorry that I came off a confrontational.

    I agree with your position. If the NDP and LPC had more seats it would be difficult for Iggy to rationalize, assuming he ruled out a coalition or working arrangement with the NDP, without first seeing what Harper has to present. Harper does get to meet the HOC and present a Throne Speech. Might be better to defeat him shortly after the Throne Speech and then hope the GG calls upon Iggy to form a government. I don't see Harper pulling a Frank Miller.

    I also can see the point you made earlier about Iggy being prepared to wait without a coalition. Weak minorities seldom last long and in all likely hood the LPC would win the ensuing election.

    Thanks for your take.

    Earl

    ReplyDelete
  26. "Weak minorities seldom last long"

    I'm not so sure about that - the Tories were in what was considered a "weak minority" position after the 2006 election and that government lasted almost three years!

    ReplyDelete
  27. DL: We had four federal by-elections in November in ridings that should have been prime territory for the Green party - they got between 2% and 4% in each of them.

    More recently, at the provincial level by-elections were held March 4th in Ottawa West--Nepean and Leeds—Grenville. The Green passed the Dipper in one and came within 35 votes in the other. Greens rose and NDP dropped in both cases.

    In Ontario, I'm not sure that most voters realize there's a difference between the federal and provincial parties. These are definitely transferable results.

    Also, could you clarify why New Westminster—Coquitlam, Cumberland—Colchester—Musquodoboit Valley, Montmagny—L'Islet—Kamouraska—Rivière-du-Loup and Hochelaga are all "prime territory for the Green party"? I have trouble assigning that label to any of them.

    ReplyDelete
  28. Volkov
    "I've been pretty clear and consistent in saying that the only way such a coalition would be legitimate is if the combined seat total was more than the Conservative seat total, and at least by 10 seats."

    Technically the total doesn't even have to exceed the CPC. That said I think more than a five seat difference is sufficient.

    ReplyDelete
  29. DL,

    The 39th Parliament is a bit of an odd duck as "weak minorities" go. The Liberals were in disarray yet still haughty about their chances, Dion never triggered a confidence vote and when he had to, his caucus fell apart, and overall the Opposition was ineffective. It lasted three years because the Liberals let it last three years.

    If a new election was called, and Harper had a weak minority, meaning under 130 seats (I think that's fair to call a "weak minority," eh?), and the Liberals had more or less 100, you wouldn't get the same result as 2006 in terms of Liberal mood and organizational state. Ignatieff would probably stay on, pull a Pearson by waiting awhile, then going off into victory as the Conservatives fall into disarray.

    ReplyDelete
  30. Peter,

    Nah, thats too opportunistic. The total should exceed the Conservatives for any legitimacy in the eyes of Canadians. It's simple math. If Lib+NDP = >Con, then its good. If Lib+NDP = Con, you're going to run into some problems.

    ReplyDelete
  31. Er, it messed up my post.

    I mean to say, if Lib+NDP = Con, you'll have some problems.

    In other words, no Bloc.

    ReplyDelete
  32. Apparently you can't do Lib+NDP(plus)Bloc on this comment form.

    ReplyDelete
  33. John don't know about the other three but the Vancouver riding is pretty granola.

    The Green vote dropped in that riding from 7.2% to 4.3% which isn't exactly a good sign.

    Then again the NDP candidate who ended up winning as a well known environmentalist who stole some of the green thunder.


    That's also a problem isn't it ? Other parties can just nominate someone in the environmental movement in strong green ridings.

    ReplyDelete
  34. Shadow: That's also a problem isn't it ? Other parties can just nominate someone in the environmental movement in strong green ridings.

    In theory. But other parties are also trying to take a slice of their Conservative/Liberal/NDP/Bloc swing vote. It's hard to appeal to four other marginal constituencies simultaneously. And besides, They're Not That Organized. (Cast your mind back to the last nomination meeting you attended and the reasons that the victorious contestant won.)

    I look forward to the day that other parties see the Green candidate as the main opponent in more than a handful of ridings. Give it a few years.

    ReplyDelete
  35. John said, "Obligatory predictability for Éric: fall election. Small Tory plurality. First Green or Greens elected. Non-confidence vote. Grit government. No coalition, but Dipper/Green/Bloc support. Stability for several years."

    John,

    I respectfully have to disagree with you. As has been mentioned before unless the NDP and Libs win enough seats to form a majority there won't be stability for several years. The Bloc are an enigma and will vote for things that only benefit Quebec. It could quite easily happen that the Libs or Dippers propose something that they don't agree with and vote down the government. So there will either be tons of Quebec vote buying (more programs etc or the will lose Bloc support). Also, I don't believe the Greens will win any seats so saying that the Libs will rely on Green support in a bit of a misnomer, IMO.

    Also, if I am wrong and the Greens do win a seat, do you think it would be Ms. May or someother seat?

    Rocky

    ReplyDelete
  36. Keep in mind that if the Liberals form a government after the next election, Harper will almost certainly quit the next day and the Tories will then be leaderless and in disarray - kind of like the Grits were after losing power in '06). We would probably end up in a situation that is the reverse of what we have now where a Liberal minority government brings in rightwing policies and the leaderless Tories abstain and let then do as they please.

    ReplyDelete
  37. Rocky: As has been mentioned before unless the NDP and Libs win enough seats to form a majority there won't be stability for several years. The Bloc are an enigma and will vote for things that only benefit Quebec. It could quite easily happen that the Libs or Dippers propose something that they don't agree with and vote down the government.

    This line of reasoning would have been compelling a few years ago, but the Bloc can apparently live with a Tory minority government who are a lot farther away from them on non-Quebec issues. A Grit minority should be at least as stable as the current situation. The actual distribution of seats would certainly affect stability and even legitimacy; Ignatieff would not move immediate non-confidence in a Conservative government returned with 152 seats. A Grit plurality or even a tiny Tory plurality would be a very different story. As always, there's a grey area in between.

    Also, if I am wrong and the Greens do win a seat, do you think it would be Ms. May or someother seat?

    My copout answer is the technically correct, "yes". There could be no Green seats or there could be several. The former is more likely than the latter today, but the longer the election is delayed, the better the Green prospects.

    Look where Greens have done best in previous elections for the most likely breakthroughs outside of SGI. Beyond that, my crystal ball requires some polishing.

    ReplyDelete
  38. John said, "Obligatory predictability for Éric: fall election. Small Tory plurality. First Green or Greens elected. Non-confidence vote. Grit government. No coalition, but Dipper/Green/Bloc support. Stability for several years."

    John,

    I respectfully have to disagree with you. As has been mentioned before unless the NDP and Libs win enough seats to form a majority there won't be stability for several years. The Bloc are an enigma and will vote for things that only benefit Quebec. It could quite easily happen that the Libs or Dippers propose something that they don't agree with and vote down the government. So there will either be tons of Quebec vote buying (more programs etc or the will lose Bloc support). Also, I don't believe the Greens will win any seats so saying that the Libs will rely on Green support in a bit of a misnomer, IMO.

    Also, if I am wrong and the Greens do win a seat, do you think it would be Ms. May or someother seat?

    DL,

    I disagree, I don't see Harper going anywhere until there is a leadership review and then who knows. He still controls his own destiny with in the Conservative party.

    Rocky

    ReplyDelete
  39. DL: Keep in mind that if the Liberals form a government after the next election, Harper will almost certainly quit the next day...

    If the Tories are crushed and the Grits have a majority or a near-majority, maybe. If the Tories have a plurality, there's no obvious reason why Harper would hand over the reins. He may not have delivered a Conservative majority yet but he's kept them running the country through two elections. Compare and contrast to Tory fortunes over the last century.

    If he loses the election after that, the knives might come out. Otherwise, it's hard to call him a failure, regardless of what you think of him.

    ReplyDelete
  40. Its not so much that the Tories would purge him - i think that the pattern with Harper is that he's a spoiled sport, a quitter and a sore loser. He supposedly came very close to quitting in a huff after losing in 2004. He also resigned from parliament in 1997 in a huff after some imagined slight from Preston Manning. Its common knowledge that Harper hated being opposition leader. I think that the moment he ceases to be PM, he will sulk and quit in a huff and give a Nixonian "you don't have Stephen Harper to kick around anymore because this is my last news conference" and then he will serve on some corporate boards and maybe run the Fraser Institute and that will be the end of him. He won't be a Diefenbaker - he will quit the moment he is out of power. Period.

    ReplyDelete
  41. This is why poll averaging is crucial to determine the true projections.

    ReplyDelete
  42. Volkov

    "If Lib+NDP = >Con, then its good. If Lib+NDP = Con, you're going to run into some problems."

    That's what I meant. You said 10 seats, I said five. Either way that exceeds the Cons.

    Now theoretically, given Bloc support that isn't necessary however that couldn't be "sold" worth a damn.

    So Lib+NDP has to exceed the Cons by at least 3-5 seats

    ReplyDelete
  43. Peter,

    My apologies then, because I thought you said that they didn't need to have more seats than the Tories to form a coalition.

    3-5 seats could cut it, yes, but 10 seats is reasonable and its safer. Plus, double-digits looks more powerful.

    ReplyDelete
  44. Volkov

    What I said was "Technically the total doesn't even have to exceed the CPC."

    Legalistically as long as they can depend on say Bloc support they don't have to exceed the CPC in seat count.

    That said it just would NOT sell at all.

    ReplyDelete
  45. Volkov that's a pretty tough goal to achieve if that's the standard.

    Given Harper's control over election timing and his risk adverse tendencies I doubt he'd go to the polls without pretty strong numbers.

    My guess is that unless he climbs back up to 40% in the polls, which is unlikely, he won't risk it.

    Legally required fall 2012 election it is!

    ReplyDelete

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.