Wednesday, April 11, 2012

Wildrose lead over PCs narrows

Wildrose is still on track to win a strong majority in Alberta's provincial election, but the margin between it and the governing Progressive Conservatives has narrowed over the past week.

Though the projection was updated yesterday afternoon to reflect the latest Léger Marketing poll, it is more helpful to compare today's update (which also includes the new Forum survey) to that of Apr. 5, when a batch of polls were added to the projection model. This allows us to contrast the changes in the aggregate rather than how the individual Léger and Forum polls have shifted the projection back and forth.

Since Apr. 5, Wildrose has slipped 0.4 points to 40.9% support. The PCs have picked up 1.2 points and now sit at 33.4%, a narrowing of the margin between the two parties by 1.6 points.

The Liberals have dropped 1.1 points to 11.1% support, while the New Democrats are up 0.4 points since Apr. 5 to 10.8%. The Alberta Party is down 0.2 points to 2.1% support.

Due to this shrinking gap, Wildrose is now projected to win 56 seats, down two since before the Easter weekend. The Progressive Conservatives are up two seats to 26, while the New Democrats are unchanged with four seats and the Liberals with one.

But the polls are beginning to agree with one another less, meaning the likely vote ranges have widened significantly. Wildrose is now projected to take between 35.9% and 45.9% of the vote, a much greater range of likely outcomes than that of Apr. 5 (39% to 43.5%). The Tories are likely to take between 30.5% and 36.3% of the vote, overlapping only a little with Wildrose.

The Liberals would take between 9.6% and 12.6% of the vote if an election were held today, while the New Democrats sit between 9.2% and 12.4%. The Alberta Party would likely receive the support of between 1.4% and 2.8% of Albertans.
This wider range of outcomes means a wider range of likely seat wins: Wildrose would win between 33 and 76 seats if an election were held today, compared to between 9 and 52 seats for the Tories. This runs the gamut between PC disaster to majority, though anyway you slice it Wildrose is poised to make historic gains. Of course, something near the projected result is the most likely, so the chances of the Progressive Conservatives forming a majority government or being reduced to single digits are quite low.

The New Democrats should win between one and seven seats (the first indication that the NDP could actually suffer losses) and the Liberals between one and three seats.

Almost all of these Liberal and NDP seats are expected to come in Edmonton, where the two parties are neck-and-neck in the battle for third place. The Liberals have the inside track, but have lost one point since Apr. 5 and now sit at 14.7% support. The New Democrats are down 1.2 points to 14.6%, while the Alberta Party is down 0.4 points to 3%. All three parties are well behind the Tories and Wildrose. The PCs are up a point to 36.9% while Wildrose is up 1.4 points to 29.2%, though their likely ranges overlap a little: 33% to 40.7% for the Tories and 24.7% to 33.7% for Wildrose.

Calgary vote projections
The best piece of news for the Progressive Conservatives, however, is in Calgary. They have gained 1.9 points and two seats since Apr. 5, though they still trail Wildrose with only 32.3% support. But it is a reversal of a slow decline in the city, as the chart to the left shows.

Wildrose is up 0.9 points to 46.3% and is projected to win 24 seats. After the blip that was caused by the Léger poll, Wildrose is back to making gains in Calgary.

The Liberals, however, are down significantly. Since Apr. 5, they have slipped 2.9 points to only 10.1% support. Note that the Liberals took  32.7% of the vote in the city in 2008. That is a remarkable collapse for the party.

The New Democrats, however, are up 0.2 points in Calgary. They are out of the race in the city, though, as they have only 7.4% support.

That betters their 6.5% support in the rest of Alberta, a drop of 1.1 points since Apr. 5. Wildrose leads here with 52.5% support, a gain of 0.9 points. The Tories are down 0.3 points to 29.9% while the Liberals are treading water, picking up 0.1 point to reach 7.1% support. The Alberta Party is up 0.3 points outside the two main cities and are projected to have the support of 2.1% of voters in this region.

The two polls that have come out so far this week are somewhat divergent, and this is the main reason why the projection ranges are currently so wide. Léger polled over the Easter weekend, and found that the margin between Wildrose and the PCs had shrunk by 6.1 points. Wildrose slipped 5.8 points since their poll of Apr. 2-4 to 35.5% while the Tories gained 0.3 points to reach 34.2% support.

Forum's poll, conducted on Monday, instead found the gap between the two parties to still be very wide: 12 points. Wildrose's support (43%) was unchanged from the Apr. 2 poll, but the Tories picked up two points to hit 31%. The numbers and margin might be quite different (as are the methodologies used, the number of Albertans surveyed, and the field dates), but these two polls do seem to indicate that the race is tightening up.

This is, perhaps, not too surprising. Wildrose stormed ahead early in the campaign, and now that Albertans are paying more attention to their policies and candidates there is a bit of blow back. Both Léger and Forum showed that some of the key campaign promises of Wildrose are not very popular, and this kind of momentum was never going to be sustainable.

I expect that the gap will continue to narrow in the coming days, making tomorrow night's debate all the more important. Danielle Smith can solidify her lead with a good performance, or she can lose that lead by allowing Alison Redford to define her and her party as unprepared for office. How Brian Mason and Raj Sherman, the NDP and Liberal leaders, will treat the two frontrunners could also be a factor. If they focus on Smith, they might push her fence-sitting voters over to the Tories. If they focus on Redford, she will have to spend part of the debate defending herself rather than going after Smith. It should be interesting to see how it plays out, and how the numbers move over the weekend.

Note: The riding-by-riding projections can be accessed by clicking on the "Alberta Riding Projections" banner at the top of this page, while full regional breakdowns can be accessed by clicking on "Regional Vote and Seat Projections" in the right-hand column. You can also check out which polls are included in the projection and their relative weight by scrolling to the bottom of this page. These features are unavailable on the mobile version of the site.

40 comments:

  1. As I suggested in an earlier thread here. This is too volatile to produce definitive projections. At the moment we're looking at trends.

    Won't see finality for about 8 days IMO

    ReplyDelete
    Replies
    1. That's probably true, but it's no reason for Eric to stop feeding our habit for a week.

      Delete
    2. Yeah, his blog is all *about* following the trend. Why would he stop just because the balls are in the air and no one knows where they'll fall?

      Delete
    3. Oh agreed completely.

      Just following the trends is fun.

      Delete
  2. Une petite suggestion pour les campagnes électorales: il serait bien de rappeler la date du vote et la composition de l'assemblée sortante à chaque update.

    Je suis la campagne albertaine uniquement sur votre site (très bon en passant), mais comme je ne connais pas beaucoup la politique de cette province, ne pas connaître ces informations de base m'empêche de me faire une bonne idée de la dynamique là-bas. Merci!

    ReplyDelete
    Replies
    1. Bonne idée, Je vais mettre cette information dans la colonne à droite.

      Delete
    2. There are a few of us in Rural Alberta we just don't comment much. By rural I mean Lethbridge and area.

      Delete
  3. I have met a large number of people who have told me they were planning to vote NDP/Liberal and are now willing to hold their nose and vote PC.

    This might be what we're seeing in Edmonton and Calgary, the collapse of the Liberal vote and a small upswing of PC vote.

    ReplyDelete
    Replies
    1. It's really a pity. A look at each of the party's responses to pressing municipal issues (at citiematter.ca) show that the Liberals clearly have the best vision for both Edmonton and Calgary. I also listened to Raj Sherman's phone-in interview on CBC and looked at each of the party's electoral platforms. I would again have to give the Liberals full credit for coming up with some creative and original policies. It is unfortunate that they have not been getting anywhere near the media coverage that I feel they deserve. I look forward to watching Thursday's debate to see if Sherman can turn a few heads and try and make this a three-way race.

      Delete
    2. I think it is a pity, but that's what happens when you're a third/fourth rank party.

      Many could have said the same of the NDP federally in years past, and many might say the same of the Liberals federally now, depending on how they feel.

      Delete
    3. Left-leaning Albertans who decide to vote Tory are doing themselves a dis-favour. The threat of a Wildrose Government is inconsequential to the continuance of a one party system that provides little accountability or choice and prevents the development of a competitive two party system.

      A supporter of the Liberals or NDP wants to see their party in government; such an outcome will only come about through a political realignment of the political system. For both these parties the best hope of that realignment would be for Wildrose to replace the Tories with a large section of former PCs (one presumes Red Tories) homeless until the next election thereby, setting up a left-right dichotomy in Alberta. Strategically voting Tory, if successful, would extend a dynasty they do not support and that makes little rational sense.

      The other thing I would like to point out is the historic ineffectiveness of "strategic voting". It rarely if ever works. I do not have any hard data unfortunately (although I am sure it exists) but, anecdotal evidence. In 2004 the Liberals made much about voting strategically to keep "Harper's hidden agenda" from 24 Sussex. It was hoped Red Tories and soft Dippers would vote Liberal. In 2004 the Liberals won a minority government and the "left" vote was one seat short of a majority (Grits+Dippers). While specualtive it is possible that strategic voting cost the NDP a seat and hence the opportunity for a Liber-NDP coalition in some form.

      The other example I would like to showcase is the 1952 BC election. The coalition government (Conservatives and Liberals) brought in Single Transferrable Vote (STV)in order to mitigate a split in the free enterprise vote (Liberals and Conservatives each had candidates in most ridings). In theory a Conservative voter would pick Liberals second and vice versa. What happened however, was many Conservatives and Liberals picked Social Credit as their second choice thereby fracturing the coalition.

      I think "strategic voting" has a dubious success rate.

      Delete
    4. In this case these are soft libs/ndp who aren't 100% satisfied with the PCs but fear the wildrose like the plague.

      You are right they are probably doing themselves a disservice.

      Delete
    5. One example of successful strategic voting comes from the 1991 Louisiana gubernatorial election when people voted for Edwin Edwards, a hugely unpopular figure and acknowledged criminal (later imprisoned on a conviction of racketeering) rather than David Duke, a leader of the KKK. Bumper stickers of he time said : I'm voting for the crook, not the Nazi. But strategic voting probably can only "predictably" work in a genuine two party system.

      Delete
    6. Might as well mention how the French leftists voted conservative rather than Nazi when LePen ranked second in the first round of the French Presidential elections.

      Delete
    7. Well, I think tonight's debate is the pivot point in the campaign. Basically, it's a make it or break it moment for the Liberals. Having seen all the platforms, the Liberal platform seems (to me at any rate) to be the most fiscally credible and there is enough wiggle room on the proposed tax hikes that would allow them to back away (if politically expedient) and still deliver a balanced budget. If Raj Sherman can deliver an ace performance and project himself as a serious alternative (which I believe he is), he has a shot at building up some vote share (a la Jack Layton in the last federal election) and wreaking some havoc.

      Delete
  4. Eric,
    My God, you're going to give me heart attack.
    Nice to see WR up again, but still too close to call. The truth is, after the election, we could still be living under the PC Jackboot.
    I simply can't wait to vote for the Wildrose.
    I think after the election I'll try to set up a support group called "308-anonomous" to help others wean themselves off your highly addictive site.

    Go Wildrose!
    40 Years is Long enough!
    Down with the PCs!

    Wildrose Supporter (Former PC Supporter)

    ReplyDelete
    Replies
    1. lol, you are right. This site is addictive. Consider me someone who will join 308-anonymous.

      Delete
  5. The Leger poll was given too much credibility. A 25 point swing with a non-random sample internet poll, only outside of Calgary and Edmonton in 4 days is not creditable. This poll should never have been considered. The latest poll shows very little change in support for the Wildrose over the last 4 days and therefore supports the previous many polls done last week. What was the spread of the last 10 polls? How does that compare with both of the Leger polls? Plot them and it is obvious who is the outlier.

    For a non-partisan poll commentator, I would expect a more balanced commentary than this.

    ReplyDelete
    Replies
    1. It looks like the outlier now, but we can't know what the next week of polling will show. In the first week of campaigning, the polls showing Wildrose with a double-digit lead looked like the outliers.

      I include all polls, whether outlier or not. Including all the polls together smooths things out.

      The margin of error in sub-regions is quite large anyway, so I wouldn't get too worked up over it.

      Delete
    2. If I could amend that slightly, Éric, you include all credible polls. You (rightly) excluded Robbins Reasrch from our federal projection.

      But Leger has an established track record as a credible pollster, so you include their polls.

      Delete
    3. Perhaps one further refinement to what Ira proposed, you include all polls from all credible pollsters. If it was a poll-by-poll determination of what was credible, it would have taken a lot longer for the model to include the Orange Wave, for example. Whereas excluding unbelievable pollsters like Robbins SEC would not cause that problem.

      Delete
    4. I shouldn't have been quite so harsh in my criticism as I am totally addicted to your site and I appreciate what you have created.

      By the way, another poll coming from ThinkHQ that also makes the Leger poll look stupid.

      Delete
  6. I bet Wildrose will get a big election day bump just like the federal Conservatives because:

    a) Voter intensity
    b) Voter demographics

    Lately Allison Redford has been putting out ads saying "this isn't your father's PC party" and counting the number of minorities/women on the WRA website and saying they're the party of "old, white men".

    Doesn't she know that old, white men vote in the highest numbers ?

    Right not she's running a pure save the furniture and hold Edmonton campaign. She's lost Calgary and she's lost rural Alberta and she knows it.

    So don't be surprised if this is still a blow out on election day.

    ReplyDelete
    Replies
    1. An election result that looks just like 1993 wouldn't surprise me at all.

      In vote totals, 1993 looks close, but the Liberals ran away with Edmonton while the Tories ran away with everywhere else.

      Delete
  7. Is your model basically a regional swing model? Are there adjustments for star candidates, incumbency etc.

    I ask because I live in Calgary Mountain View, and based on lawn sign counts David Swann (L, incumbent and former leader and fairly popular in the riding as far as I can tell) might even be ahead, while your projection shows the Liberals more likely to finish in 3rd. Certainly anything except an NDP win could happen in this riding, but I suspect Lib/Wild Rose will finish first and second in some order.

    ReplyDelete
    Replies
    1. The model is a regional swing model that makes adjustments for incumbency, etc. Full explanation of the methodology can be read by clicking on the "Seat Projection Methodology" link in the right-hand column.

      Individual ridings can always buck the trend, but in this case take a look at the ranges. Swann could get as much as 30% of the vote based on the current projection, which puts him in the running.

      It is just difficult for the Liberals in Calgary when they have lost 2/3rds of their support, but certainly there are always exceptions.

      Delete
    2. As a former Calgarian, I'll concede that if any Liberal can get elected there it's probably David Swann. He was (if I recall correctly) the founder of Friends of Medicare, the group that inspired Ralph Klein to create his Truth Squads (really, that's what they were called).

      Delete
    3. David Swann will lose his riding. I am actively trying to make sure of that. He's got lot's of signs, but there is a quite wave of support.
      Wish I could say more.....

      Anyway, the only Liberal that I think can and should win is Kent Hehr in Calgary Buffalo. He has deep roots in the community and unlike Swann, he's effective and smart.
      The thing with Hehr is that his riding is jam packed with enormous highrises and multi-unit buildings. This makes the turnout unpredictable, since you have so much trouble getting data and counting lawn signs.....
      We'll see.
      Loving the individual riding polls Eric!
      (A little too much...I a total nerd)

      Delete
  8. The Wildrose party has muzzled its candidates - scary having Danielle Smith as the sole spokesperson. She is unfit for elected office of any kind.

    ReplyDelete
    Replies
    1. How, exactly, is Danielle Smith scary?

      Delete
    2. That is actually more misinformation. The candidates have to talk to the Wildrose war room prior to major interviews, but there is no muzzling. You are confusing her with the PCs that kick candidates out of party if they don't tow the party line.

      I

      Delete
    3. Danielle Smith.
      Will put Soldiers.
      In our Streets.
      We are not making this up.

      Down with the evil PCs!
      40 years is long enough!
      Long live the revolution!

      Wildrose Supporter (Former PC Supporter)

      Delete
    4. I actually think the first thing that has made me rethink voting PC was saying "this isn'y your fathers PC party".

      My dad is an extremly successful man who supported his wife so she could stay home and raise her litter of children. If I'm half as successful in my life as he is it will only be through hard work and dedication.

      Maybe Allison Redford has been taught since birth to hate her dad for being an evil white man, but all that shows is her (and her handlers) projection.

      Delete
    5. Sounds like left wing fear mongoring, no substance.

      Delete
  9. http://quebec.huffingtonpost.ca/2012/04/11/sondage-quebec-intentions-vote_n_1419215.html?ref=fb&src=sp&comm_ref=false#sb=2727999,b=facebook
    Sondage: la CAQ s'effondre, le PLQ remonte, le PQ bien en selle
    Le Huffington Post Québec | Par Eric Grenier 12/04/2012

    ReplyDelete
  10. 1 in 20 outlier13 April, 2012 11:42

    Ipso Reid came out with a poll that will be used to establish the debate as the great turning point in the election.

    http://www.qr77.com/News/Local/Story.aspx?ID=1685879

    Smith won the debate outright 37-28 but improved her position by 10% and Redfern lost 15% on her position,

    The Turning point actually being Redfern and the unions beating out Morton for the leadership..... or further back Dec 2,2006 when Stelmach jammed the ballot box to beat Dinning for the leadership.

    It will be interesting who comes in second...

    With the WR going to form a majority it might be better for the progressives supporting the PCs to vote NDP.

    The NDP leader in the debate sounded a lot like small "c" conservative attacking Redfern for overspending money she didn't have.

    ReplyDelete
  11. I took Eric's seat projection and plotted the difference between the PC and WR popular vote in each riding. I counted all the ridings where each party was projected to win by 8% or more. All the seats that could be won by the ND, Liberals, or other were not counted. My "safe seat wins" came to 42 for WR and 17 for PC. A majority is 44 so the WR would only have to win 2 seats out of 28 (less than 8% difference) to get a majority government.

    Sure looks like a Wildrose majority government to me!

    ReplyDelete
  12. I just saw a new poll by "Return on Insight" for the CBC that puts the WR ahead by 21 points in rural areas (about the same as your prediction) but in Calgary they have the WR down by 4 (WR 41 PC 46) and in Edmonton the WR ahead by 6 (WR 37 PC 31). This poll is as of April 16.

    Is it possible that the Calgary WR vote could go from 17% up to 6% down since April 10? Could the Edmonton vote go from 11% down to 6% up?

    The pundits say that the debate changed nothing and I haven't seen any articles that would explain these numbers. Do you know something that might help?

    ReplyDelete
    Replies
    1. Until we see some other evidence that Edmonton and Calgary have flipped, we might just have to chalk it up to small samples.

      Delete
  13. Hi, I live in Calgary mountain view and want to vote for the candidate most likely to defeat Wildrose. Should I be voting for Swann or the PC candidate.

    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.