Wednesday, December 11, 2013

November 2013 federal polling averages

It was a much quieter month on the polling front, with only four national and regional polls being released throughout November, down from 14 in October. The polls that were out showed general stability, if not an uptick in Liberal support. But due to the lower sample size, I am just going to briefly go over the averages in November for the sake of continuity.

Kickstarter Update: We're now at 58% funding at the 50% mark of the pledge drive - 15 days to go! The project currently has 123 backers. Thanks so much! Let's keep spreading the word.
The Liberals averaged 36.5% support in November, up 1.3 points from their October average. The Conservatives were down one point to 27.9%, their lowest since May and their second worst result since at least before 2009.

The New Democrats were down just 0.4 points to 23.4%, while the Bloc Québécois was up 0.5 points to 6.4%. The Greens were down 0.6 points to 4.9%. Support for other parties stood at 1%.

If we compare the national polls that were in the field in November to the last time these firms were last active, we see that the apples-to-apples comparison is even more favourable to the Liberals.
Regionally, the Liberals led in British Columbia, Ontario, Quebec, and Atlantic Canada, while the Conservatives were ahead in Alberta and the Prairies. The NDP was second in British Columbia and Quebec.

There were a few notable regional results, however. The Liberals scored their best-ever (at least since January 2009) results in British Columbia and the Prairies, while the Conservatives managed their worst results since before then in Ontario. On the other hand, they had their best score in Alberta since February 2013, while the Liberals were at their lowest point since then. The NDP had their best result in Alberta since October 2012, but dropped in support for the third consecutive month in Quebec.
With these numbers, the Liberals would likely win 142 seats on the new 338-seat electoral map, with 117 going to the Conservatives and 68 to the New Democrats. The Bloc would win 10 and the Greens would keep their one seat.

This is a gain of 12 seats for the Liberals since last month. The Bloc dropped five, the Conservatives dropped four, the NDP dropped two, and the Greens dropped one.

The most important changes took place in Ontario and Quebec. The Liberals and Conservatives swapped seven seats compared to the October projection, while the Liberals picked up nine in Quebec. Four of them came from the NDP and five from the Bloc.

So, another month of the Liberal lead since Justin Trudeau became leader. It now stretches to eight months, equal to the amount of time the Conservatives led in the polls after the NDP relinquished the lead in 2012. This Liberal lead is the longest the party has enjoyed since before the January 2006 election, and by far the most significant in size since then. This just may be the new normal for the time being.

35 comments:

  1. So I guess everyone to said this was just a honeymoon is eating crow now.

    ReplyDelete
  2. Eric I'm a little confused by one thing. You show Alberta as a separate and then a "Prairies" thing. But I always thought Alberta was part of the Prairies ??

    Or by Prairies do you only mean Sask and Man ??

    Maybe we should start looking at all provinces and territories individually ???

    ReplyDelete
    Replies
    1. The Prairies are Saskatchewan and Manitoba, as that is how virtually every polling firm divides up the country. It is impossible to look at every province individually, as the polling data for the Prairies and Atlantic Canada does not exist.

      Delete
    2. "It is impossible to look at every province individually, as the polling data for the Prairies and Atlantic Canada does not exist."

      Then I would suggest some polling firm get their act together. Atlantic Canada = FOUR provinces, that makes the result twice as bad as the "Prairie" result !!

      Delete
    3. Peter, the problem is the sample sizes. You'll note that we don't even have data for the three territories, an even worse situation than lumping together the four Atlantic provinces. If you wanted the polling firms to break it down to each province, you'd (realistically) end up with a sample for PEI in the neighbourhood of, what, 15 people? That wouldn't be very meaningful. Also, how significant is it to have a close reading of the 4 seats in PEI vs. a comparatively distant reading of the 106 in Ontario? The principle is noble, but are resources best used to poll by province or by regional population of comparable size?

      Delete
    4. You can't lump together disparate entities chim !!

      And that's exactly what's happening here right now. Take the prairies, two provinces with different provincial govt's and you lump it all together !! Same with Atlantic Canada !!

      Territories are a different issue but one that needs to be explored !

      Delete
    5. You could make the argument that it makes no sense to clump together the West Island with the Saguenay in Quebec, or Vancouver with northern BC, or northern Ontario with Toronto, etc. etc.

      It is done this way because of sampling issues, always has been, always will be. Best come to peace with it.

      Delete
    6. But it's only a question of scale. By the same logic, you can't lump together separate entities on a provincial level either. I live in a downtown Montreal riding that has been Liberal since the beginning of time, but it gets lumped in with all the other Quebec ridings for the purposes of federal polls (and we all know that until Justin Trudeau came on the scene, the NDP had been the colour of Quebec since the 2011 election).

      Delete
    7. You've just proven my point chim, thanks.

      Apparently unless the population or economic activity exceed a certain point pollster feel quite satisfied to ignore realities, eh?

      Delete
    8. If someone is prepared to pony up the dough for sample sizes in the hundreds of thousands, I'm with you. Otherwise, provincial divisions are as arbitrary as any other divisions you care to name...The only non-arbitrary divisions are those by riding, since those are the units actually being measured. That's what I meant by resources, I wasn't referring to the economic significance of the region being polled. Why should the four ridings of PEI count more than any other four ridings in the country?

      Delete
    9. A lot of media is done provincially, and the common provincial government binds economic interests together to an extent.

      Delete
    10. YesRyan but that doesn't excuse putting two provinces into one as the "Prairies" or four into one as in Atlantic Canada !

      Delete
    11. Sample sizes. Smaller and smaller samples mean more and more meaningless numbers.

      Delete
    12. If anything, the main issue here being sample size, the main problem is not lumping Atlantic Provinces and Saskatoba together, but not regularly compiling data for urban centers. Montreal, Toronto, the Lower mainland in BC are all large and distinct enough to constitute meaningful divisions in national polls. I'd say urban, suburban, and rural Canada are also large enough and meaningful enough categories that are being regularly ignored by polling companies but could easily be included in any analysis.

      Delete
    13. Assuming an average riding size of 150,000 people, a margin of error of +/- 3% and a 95% confidence internal, it would take 500 calls per minute to conduct a national survey within 12 hours. Rather than every poll covering every region equally, different polls by different polling agencies vary their polling areas. For example, EKOS splits up Manitoba and Saskatchewan, Harris Decima has the two together, but both report Atlantic Canada as a group, while Ipos Reid and Forum Research only report national results. This site therefore has to either lump the areas together to have more applicable data or separate them and have less counterbalance. I'm happy with the way things work and the site's predictions show they are reliable groupings.

      Delete
    14. Ipsos and Forum report regional results as well, with Saskatchewan and Manitoba lumped together. The only firms who do not do that are EKOS (which separates them into almost useless samples) and Nanos (which includes Alberta into the mix, or at least did back when they reported voting intentions).

      Delete
    15. Incidentally, don't you find the Nanos index terribly frustrating and borderline useless? You can't compare it with any other polling and you don't really know how the idnex is generated... It seems to function, more than anything, as a fairly arbitrary scale with little accountability... Unless I'm missing something?

      Delete
    16. Interesting to compare how the different polling firms have adjusted their practices following the Alberta, BC, etc. elections...

      Delete
    17. Indeed. Angus-Reid, for example, hasn't released a voting intentions poll since the BC election.

      Delete
    18. well, I suppose hiding is one option...

      Delete
    19. Would you consider a full full review of the various polling firms' responses to the 'errant' elections?

      Delete
  3. Eric, do you have any explanation for the large jump in PQ support in the CROP poll and the large jump in satisfaction with the PQ government in the CROP poll. Both of those things contradict the Leger poll released just a few days earlier and nothing has happened in the days between when the polls were conducted to change things. Also, the jump in support recorded in CROP for the PQ was not at all reflected in the by-election where the PQ ran a candidate. In the Viau by-election the PQ dropped around 9 percentage points from the 2012 election. So what is going on here?

    ReplyDelete
    Replies
    1. There hasn't really been a big jump in support in the CROP poll, the PQ went from 32% to 35%, which would be within the margin of error of a probabilistic sample. So it could just be a wobble.

      As for the increase in satisfaction, that is a bit of a puzzler. It could be exaggerated a bit, as the last poll had them down to 32% on satisfaction from the usual 35%. But it is notable that most of the gain came from those saying they were 'mostly satisfied', so it might not take much for people to go from mostly dissatisfied to mostly satisfied. The feeling isn't particularly strong.

      The poll is not at odds with Léger in terms of support (again, it might just be a wobble) but is in terms of satisfaction. Could just be the vagaries of probability, here.

      And for the by-election in Viau, turnout was so low (just 17%!) that I wouldn't read much into it.

      Delete
  4. I think the new normal comment might be the most accurate. It sure seems like Trudeau has given the liberal party the upswing and held on to it in the way the party had wanted. Its reasonable to think that in order to come back from their 3rd place finish in last election they (the party) needed to shake it up a bit and go for something different in a leader who could connect better with people....ie different than Dion and Ignatieff, who were academics but dry toast. Difficult to make the case that Trudeau falls into that mold.

    Its quite possible that the current troubles for the conservatives will pass when it comes election time. Platforms will be released parties will attempt to run on ideas but the connection of the voting public to the leaders likely matters more and numbers always move during election campaigns. Shifted by leaders debate, media focus on who has momentum, candidates saying horrible things about rape being gods plan (er wait wrong country) etc

    ReplyDelete
  5. Just a suggestion on your seat predictions based on the poll results:

    Incorporate the amount of money the parties will have to spend on the election into the formula.

    Somehow in this new normal of Liberal popularity they are raising significantly less than the Cons --

    The first 9 months of 2013 the Liberals raised 6.8M compared to the Cons 12.7M


    Since the last election (starting Q3,2011) the Cons raised 37.2 M and the Liberals 19.2 M

    That disparity in campaign funding obviously will impact the Liberal election campaign and GOTV campaigns.

    The funding disparity indicates more than any poll that the Country is not fed up with Harper to the point that they will take action,

    It also shows that by not having any discernible platform is really hurting the Liberals ability to raise cash.

    Until the Liberals can figure out what they want to do and announce it as a goal worth giving some cash to they will be lamenting that they were totally out-campaigned in 2015 and couldn't compete with the Cons political machine.

    Why does Trudeau want to be PM? What will he do as PM?

    Right now he is afraid to answer would hurt his poll numbers..... BUT failure to answer provides no motivation to contribute to his cause.


    ReplyDelete
    Replies
    1. Yes... waiting until your party has its policy convention to release policies is a bad thing BCVOR. Give me a break. Liberal fundraising is trending up. Conservative fundraising is trending down. In fact, Liberals actually have more donors than the Conservatives now. Go figure.

      Delete
    2. Not entirely sure who has more money wins elections though. Is there any Canadian data on that?...assuming not since caps on spending seems to limit what could be accomplished...as per the american model.

      That said though, I think its safe to assume that party spending if it does make a difference is already in the polling information. People see the adds it either changes their mind or it does not. Polls changing over time during a campaign would be reflections of that. So size of the pocket book per say and if spending it does have an effect it would be reported over time

      Delete
    3. @ryan since the last election the Cons have raised $19M more. In Canada that is a huge advantage.

      It is not supposed to be a hard Question--- but it kept Ted Kennedy from becoming president--- why do you want to be PM? Of course if you don't know or won't tell it will be hard to get folks to write even a modest cheque to support your "vision".

      As for the Liberals improving fund raising --- 2009-Q2 the Liberals endorsed Ignatieff as Leader and raised 3.8M within 100,000 of what the cons raised.

      The best Trudeau has done is 2.9 2013 Q2.

      Correct me if I am wrong but it seems that Ignatieff had the Liberal base more fired up than Trudeau.

      Delete
    4. So you're cherry picking just one quarter.

      Ignatieff also had a much larger caucus to do fundraising for him.

      Also FYI of that $19M advantange, $9 million went into a now scrapped voter ID system.

      Delete
    5. Ryan

      BCVOR needs to check this out !! Tories slipping !!

      http://www.cbc.ca/news/politics/conservative-brand-slipping-in-key-regions-nanos-finds-1.2460453

      Delete
    6. The Liberal Party currently has a donation counter for December on their website (https://www.liberal.ca/million-dollars-for-change/) and are close to $1M in 13 days. The $1M target by Dec 12 was positioned as the required amount to out fund raise the CPC in the latest quarter.

      Delete
  6. While the sucess the Liberals had in 2009 hasn't been matched yet, the Liberals have had more donors than the Conservatives in the last 2 quarters for the first time probably since the Conservatives merged. The Conservatives raised more per person so they still kept ahead. I suspect that part of this was the large Conservative convention fees at over $500 per person that brought average Conservative per donor up. We'll see whether the Liberals can translate the large number of new donors that they have attracted this year to become larger donors and finally allow the Liberals to match Conservative funding. Will the use of party funds for legal fees, the cast off new data base and the willingness to pay Duffy's expenses cause a drop off in Conservative funding from the base?

    ReplyDelete
  7. And with these poll results please note NO party gets a majority Govt !!

    ReplyDelete
  8. And Paille resigns citing health problems !!

    Will this have any affect ??

    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.