Friday, September 16, 2011

Wildrose may not be out of it after all

Earlier this month, I noticed mention of an Alberta poll in an article from the Edmonton Journal. It was a poll done privately, but the results were provided to Graham Thompson of that newspaper.

ThinkHQ Public Affairs did the poll, a firm I am not familiar with but is apparently run by someone who used to work with Ipsos-Reid. It was taken in early July, so it's quite old, but it is noteworthy nevertheless as we don't get many polls from the province.
The poll pegs the Progressive Conservatives at 39% support, compared to 30% for the Wildrose Party, 14% for the New Democrats, 11% for the Liberals, and 4% for the Alberta Party.

These results are consistent with an Abingdon Research poll that was released in May, but is at odds with an Environics poll that was conducted in the second half of July. While support for the Liberals and NDP is generally the same, Environics had the PCs at 54% and Wildrose at 16%. Clearly this is a very different result, but aside from ThinkHQ being a newcomer to the political field there isn't much to doubt about this poll. With the permutations of the PC leadership campaign, it is understandable that polls could differ to this extent.

This nine-point gap isn't very useful to Wildrose, as the Progressive Conservatives would still win a large majority with 63 seats. Wildrose becomes the Official Opposition with 18 seats, while the NDP wins six. The Liberals, who at the time of this poll were still leaderless, are booted from the Legislative Assembly.

Obviously, with a new leader at the helm of the Alberta Liberals and a new leader to be chosen by the Progressive Conservatives in October/November, things are bound to change. But this poll does give us an indication that things aren't as smooth for the PCs as many may have thought. This makes the choice of their next leader very important - Wildrose is still well-positioned to take advantage of any misstep.

24 comments:

  1. The colour you use for Wildrose looks almost identical to the Liberal red. What colour do they tend to use on their posters etc...?

    ReplyDelete
  2. Progressives in Alberta should unite to support the NDP as this would make it a real competitive alternative to the other two centre right and right wing parties.

    ReplyDelete
  3. DL,

    The Liberals are red and Wildrose is a forest green in this graph. Are you, perhaps, colour blind?

    ReplyDelete
  4. Something strange has happened to Canadian political standards if anyone seriously considers shifting from fringe status to official opposition within a single election cycle with only 9% less popular votes than the government to be a result that "isn't very useful to Wildrose." Especially since that's a theoretical result, and there's still a campaign around the corner where they could close that gap.

    I don't know much about Wild Rose generally, but if I was running the Tory campaign with a new leader coming on board, I wouldn't be exactly be standing up and cheering at a nine point lead.

    ReplyDelete
  5. Yep, a useless, useless result for Wildrose ad nothing really at all. Disband already!

    Or, as I wrote at the end of my post, "Wildrose is still well-positioned to take advantage of any misstep [by the PCs]."

    ReplyDelete
  6. I think there's only two sure bets for the next Alberta legislature: the PCs will continue as the Government of Alberta, and the Wildrose Party will be Official Opposition.

    Actually, three sure bets: the next Premier of Alberta will not be Ted Morton... :)

    @Joseph: they've already done that. It's called the Alberta Party, but I don't think it'll win a single seat in the next election.

    ReplyDelete
  7. If indeed the Liberals are shut out of the Alberta Legislature, does anyone think they would collapse altogether? How likely is it that the NDP would rise to replace them as the single party on the left.
    Alberta is probably the best place to test out a working coalition between the Libs and NDP. They don't have to merge, they just need to cooperate and not run candidates against each other. Although I think the NDP would rather see the Liberals collapse than work together.

    ReplyDelete
  8. All results are meaningless until the PCs have their new result....even then, that leader will have been in the limelight for 3 or 4 months during the leadership campaign. Once the general election starts, and Albertans compare Danielle Smith with her "pull another rabbit out of the hat" PC opponent, watch for a big shift.....excited to see it happen!

    ReplyDelete
  9. The fact that the PCs will have a new leader, and that the Liberals will be led by Raj Sherman definitely makes this hard to predict, and makes a July poll not so useful in the long run, though it is interesting to see; especially the level of Wildrose support, which was commonly thought to be dropping precipitously due to Stelmach's departure satiating the right wingers.

    There are a lot of wildcards in the AB politics; the new PC leader, Raj Sherman leading the Liberals, new popularity of the NDP, the uncertainty of Wildrose success, and whether or not the Alberta Party can raise their profile enough during the campaign and in a few ridings to elect a couple or few MLAs. It's damn interesting times.

    ReplyDelete
  10. the PCs have been mis-stepping since 1972 and no one (yet) has taken advantage of it. the analyis on the surface seems reasonable. i felt that the wildrose drop in numbers was somewhat manufactured via the media's obsessive coverage of the pc leadership race.

    ReplyDelete
  11. Anon,

    Clearly you haven't studied Alberta politics. From the 1970's to 1990's, the Alberta Liberals were essentially shut out of the legislature.

    Just like how in Ontario, they've been the third party before.

    Or how in Saskatchewan and Manitoba, being shut out or down to lows such as 5% and 10% aren't new experiences.

    People should realize that the Liberals have already survived one apocalypse. We'll survive another.

    ReplyDelete
  12. Actually I do suffer from some degree of red/green colour blindness!

    ReplyDelete
  13. That must be the culprit, as to my eyes the colours I've used for the Liberals and Wildrose are completely different. Sorry I can't accommodate!

    ReplyDelete
  14. DL -- Stay away from duct tape!!!

    South Parkdale Jack

    ReplyDelete
  15. BC Voice of Reason17 September, 2011 16:26

    Eric ... your projection of 18 wild rose seats on 30% of the vote is not worth the paper it isn't printed on.


    There is no history or break down in polls that would allow you to come up with that number.

    The Wild rose vote in 2008 election was 6.8%

    If the 30% Wild rose vote is concentrated in Edmonton, Calgary or rural Alberta they could sweep any particular region and get 25-30 seats.


    when a party goes from 7% to 30% in polls there is not enough information to make any accurate seat projection.

    Your May 1 seat projections for the NDP in Quebec was 31 (they got 59)

    ReplyDelete
  16. You're wrong. With the NDP at 43% and the other parties at their actual level of support, the model would have projected 60 seats for the NDP in Quebec. The problem was that I didn't have the NDP at 43%.

    ReplyDelete
  17. BC Voice of Reason18 September, 2011 11:17

    So you were able use your model to predict the result after the election took place.

    Your model is able to predict the past!!

    And you claim success?

    ReplyDelete
  18. Sigh.

    You said:

    "when a party goes from 7% to 30% in polls there is not enough information to make any accurate seat projection."

    That's wrong. The NDP went from 12% to 43% in Quebec, and that was enough information to project the NDP at 60 seats in Quebec.

    You then said:

    "Your May 1 seat projections for the NDP in Quebec was 31 (they got 59)"

    The problem was that I had the wrong popular vote input into the model. It had nothing to do with the party going from a very low level of support to a high level of support, which you claim is a problem for projecting Wildrose.

    ReplyDelete
  19. Eric,

    Quick question: any projection system I saw had the Liberals losing all but two seats (Dion's and Pacetti's) when they fell down to just 14.7%. Obviously though, they managed to keep seven.

    So, how would your system account for something like that? Basically, how would it account for incumbent "staying power," as it were? I know you can tweak individual ridings just so to give local candidates a boost, but this seemed to be something that was a little widespread for the Liberals in the majority of ridings, but especially in Quebec.

    ReplyDelete
  20. My model would have had them at five seats with that level support (and with everyone else where they were). With the actual provincial vote numbers, the 2011 projection for Quebec would have been 60 NDP, 6 CPC, 5 LPC, 4 BQ.

    Incumbents get a bonus when their party has lost support. In the 2011 election, that bonus was 10% (not percentage points).

    ReplyDelete
  21. Ah, well fair enough! Thanks Eric.

    ReplyDelete
  22. Sorry to hijack the Alberta thread, but reports have come out in the London Free Press and the Toronto Sun of a new provincial poll in Ontario with the results PCs 36%, Libs 33%, NDP 29%, which leaves at most 1% for the Greens. These numbers are within the margin of the Angus-Reid numbers released on Saturday. No regional breakdowns or anything, but an exceptional number for the NDP.

    http://www.lfpress.com/news/ontarioelection/2011/09/18/18705281.html

    ReplyDelete
  23. I just want to say that I think Eric's model works well under normal circumstances as it filters out the "background noise" inherent in polling. But it doesn't work well when there is a sudden dramatic real shift in voter intentions during an election (like the last federal election in Quebec).

    I do tend to agree that Wild Rose results may be quite unpredictable as regional shifts may not reflect the past election. With the NDP and Liberals, the regional distribution of votes tends to follow longer term trends, so it should be easier to predict which seats might go to the NDP with its apparent rising polling numbers.

    ReplyDelete
  24. Anonymous 17:18,

    I have made some changes to the model so that it would operate more accurately in a campaign liked the 2011 federal campaign.

    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.