Friday, April 8, 2011

Conservatives back to a minority

Before getting to this morning's projection, a word on my update schedule. Today's update is being done later than usual, and much later than I had hoped I would be able to do when the campaign had started. I was aiming for updates before 8:00 AM, but that was before Nanos's reporting schedule became known. I want to include the most recent polls in my update, which means I am only starting on the projection update after 7:00 AM each morning. On certain days, like today with EKOS reporting shortly after 8:00 AM, that pushes my update back past 9:00 AM. I will try to keep the updates coming as early as possible each day, but it is difficult to keep to a consistent schedule.

But this morning's update may be worth the wait, because we have a few noteworthy changes taking place. Most importantly, the Conservatives have moved back into minority territory, thanks to new polls from the researchers: EKOS, Nanos, and Forum.
The Conservatives remain stable at 38.6% nationally, a level of support they have maintained for days. But they have dropped two seats and are now projected to win 153 - two short of a majority. The Liberals, on the other hand, are up 0.2 points to 27.8% and one seat to 72.

The New Democrats are down 0.1 points to 16.9% but are up one seat to 33, while the Bloc Québécois is down to 9.3% nationally (but still 50 seats) and the Greens to 6.3%.

Aside from a 0.9-point gain in British Columbia, the Conservatives have remained stable throughout the country.

The Liberals, on the other hand, have dropped 0.5 points in British Columbia but have gained that amount in Ontario - a significant jump. They are also up 1.1 points in Atlantic Canada, closing the gap on the Tories.

The NDP is up in Alberta, the Prairies, and Quebec, but down in Ontario and Atlantic Canada. A mixed bag.

The Bloc is down again in Quebec by 0.3 points. They are now projected to have 37.5% support.

The two seat changes have taken place in Manitoba and Ontario. The riding of Elmwood - Transcona has transferred back to the NDP from the Conservatives in the projection, and incumbent Jim Maloway is favoured again. In Ontario, Kingston and the Islands has moved back into the Liberal fold from the Conservatives. New candidate Ted Hsu is projected to win the riding for the Liberals after the retirement of Speaker Peter Milliken.

With the Liberal gain in Ontario, a few ridings are getting close to flipping to their side. We shall see where the weekend takes us, and you can read tomorrow's projection in Le Devoir and Sunday's on The Globe and Mail's website.

24 comments:

  1. 'Morning Eric.

    I'm betting the crime bill announcement today will give the Tories another boost. Do you agree?

    ReplyDelete
  2. I'm not sure individual policy announcements have that much of an effect. They certainly haven't so far.

    And on this particular issue, I don't think it is the kind of "fence sitter" issue that would bring people from the Liberals or NDP to the Tories.

    ReplyDelete
  3. Goaltender Interference08 April, 2011 10:43

    Precisely, Eric. The Tories have announced crime bills literally dozens of times over the past five years. This isn't designed to get new voters, it's designed to get those of their current supporters who are motivated by this issue to show up to the polls. There's even a slight potential to drive away a few fence-sitters, as one of Ignatieff's lines has been criticizing the Tories for "wasting money on jails and fighter jets instead of [health/education/etc.]" But like I said, Tories supporting crime bills is no surprise to anyone, so I can't imagine it being a big vote-changer either way.

    ReplyDelete
  4. It did seem to me that Harper's specific announcement supporting the hydro project for Atlantic Canada sparked a sharp uptick in the region -- but perhaps I've got the dates wrong. Anyway, the platform is out now, so we'll see.

    ReplyDelete
  5. @Shawn, I think the Conservatives being for a "law and order" platform will surprise no one, it's already well-known. They've squeezed the orange so much with all the crime bills that I don't see much more juice being obtained from it, people who would support them for that already do support them.

    Unless of course something were to happen that would bring the spotlight back to this issue. It's always possible, but it doesn't seem likely.

    What I find interesting is that the Conservatives even present a platform. They didn't in 2008, they limited themselves to criticizing the Liberals and personalizing the debate. At the start of the campaign, they seemed to me to want to do it again. They had certainly set the stage for it with aggressive ads against Ignatieff, and the few policy announcements had been very, very small and very, very far.

    So it makes me wonder if the Conservative campaign has just scrapped its initial plan and decided to go in an altogether other direction after they noticed the Liberals seemed to have momentum and not them. And after they noticed that Ignatieff didn't do the same mistakes as Dion did, namely going for policies that might make a lot of sense and do a lot of good but that would be easy to attack politically (ex: the carbon tax, maybe a very good idea for the environment, but promising new taxes isn't the way elections are won, you know what I mean?). Dion ultimately was an academician used to rational debates, which doesn't help win public debates it seems. Ignatieff is an academician too, but he always seemed to have a knack for PR, that Dion didn't have.

    ReplyDelete
  6. Nanos does show that policy of a party is the single largest reason they vote for a particular candidate 52%. Seems reasonable to think individual announcements could pay off in the longer run rather than over 1 day polling.

    ReplyDelete
  7. Tories leaked the crime omnibus strategy last night to hurt the Liberals.

    Ignatieff is under fire for trying to claim he'll save money by not building prisons.

    However, his public safety critic Mark Holland said the Liberals would NOT repeal any crime bills.

    Then Ignatieff said well maybe they would but he won't say which ones.


    How do you save money on prisons if you're not going to repeal the tough on crime legislation you helped pass ??

    Reporters just can't seem to get a straight answer out of Ignatieff on this.

    ReplyDelete
  8. If anything, the so-called "crime bill" will hurt the tories. We all know crime rates are dramatically down over the past several decades. This issue has no traction whatsoever and frankly, I doubt it even fires up their base very much. Yet another go-nowhere policy from the Conservatives. Worse still, there is absolutely no public hunger for more billions to be spent on new mega-prisons, either on the left or right. Not really sure why Harper continues to swing this dead cat around...

    ReplyDelete
  9. It seems like the only thing we know more or less for sure from the polls is that the Conservatives are in first place, the Liberals in second and the NDP in third - beyond that its a mess. Depending on what poll you look at the NDP could be anywhere from 12% to 21% in Ontario and anywhere from 13% to 28% in Atlantic. The Tories could be anywhere from 34% to 48% in BC!!

    Eric, do you know what methodology the Forum poll used? I doesn't seem to say anywhere. i know their last poll used the same IVR method that Ekos uses - but their poll this time seems to ask a lot more questions and i was under the impression that a computer can only keep people on the phone through about 5 or 6 questions before they all hang up - so I wonder if this Forum poll was IVR or traditional telephone?

    ReplyDelete
  10. Eric, I am very curious as to the weighting Nanos receives in your model. I have noted that their regional daily tracking has the GPC at under 5% in British Columbia. I know the M.O.E. is pretty large for the small regional samples, but this has been pretty consistent.
    I strongly doubt the accuracy of this number, as the GPC has polled, and voted well above this level, even before the Advent of Jim Harris in 2004 election. I find it really hard to swallow that there is no core support for the Greens in BC, which tends to cast doubt on their (Nanos) methodology.
    Also, if they are in fact in error, and the GPC will have a significant impact on electoral outcomes in 70-80 Electoral districts in Canada, I would be concerned about how it impacts your models outcomes for the Liberal and NDP seat counts. I know it is hard for pollsters to accurately predict GPC support levels, due to the historical pattern of polled support levels evaporating on EDay, but could it be that they are applying 'corrections' where it is not meaningful?

    ReplyDelete
  11. If anything, the policy announcements so far have served to frame the issues. Every time the CPC announces some program they'll expand "when we can afford it" it raises the issue of the economy in the minds of the voters.

    That's their goal, and it appears to be working.

    That said, I don't see how the crime policy fits into that. What they need now to shore up support in BC to have another boat full of Tamils arrive.

    ReplyDelete
  12. DL,

    Though I don't know for sure, generally polling firms will use the same methodologies from one poll to the next, and the large sample size tells me they were using the IVR system rather than flesh-and-blood operators.

    ReplyDelete
  13. Bluegreenblogger,

    I'm not sure what your question is. Only Nanos polls that don't overlap are included in the model.

    For example, today's poll will be replaced in the model by tomorrow's poll, and so on. Three days from now, today's poll will be returned to the model and the new Nanos poll that includes three independent days of polling will be added.

    ReplyDelete
  14. "And after they noticed that Ignatieff didn't do the same mistakes as Dion did, namely going for policies that might make a lot of sense and do a lot of good but that would be easy to attack politically (ex: the carbon tax, maybe a very good idea for the environment, but promising new taxes isn't the way elections are won, you know what I mean?)."

    But the Liberals DO have a plan for a carbon tax hidden on page 46 of their platform statement. Of course they want it to stay there, well hidden.

    ReplyDelete
  15. "We all know crime rates are dramatically down over the past several decades."

    Try telling that to anyone living in any of Canada's big cities where we dodge gunfire from drug gangs.

    ReplyDelete
  16. DL said: "Depending on what poll you look at the NDP could be anywhere from 12% to 21% in Ontario and anywhere from 13% to 28% in Atlantic. The Tories could be anywhere from 34% to 48% in BC!!"

    _____________________________

    I would take the median of your figures to get a more accurate numerical snapshot for any particular region.

    In fact, Eric's model/numbers for BC, as an example, makes absolute perfect sense. I've previously stated on here over one year ago that the template for BC is as follows: CPC - 40%, Liberal - 25%, NDP - 25%, Green - 10% and Eric's numbers currently fully reflect that.

    By the final weekend of the campaign, we will see where the momentum goes in BC, for example, and that could mean up to another bonus of ~5% for one party and another ~5% penalty for another/others.

    ReplyDelete
  17. P,

    I live in a big Canadian city and I've never been shot at. Am I doing something wrong?

    ReplyDelete
  18. P said:
    "Try telling that to anyone living in any of Canada's big cities where we dodge gunfire from drug gangs."

    I came to Canada from Bosnia 15 years ago. I live in Vancouver now, in the past 15 years I'd lived in Toronto, Ottawa, St. John's and Montreal (yep, every 2-3 years I got the itch and moved).

    I've never seen a gun in any city and I worked many different jobs.

    P, I wonder where you live so I don't move there :)

    ReplyDelete
  19. Eric,

    I was just trying to make the point that gang crime is rife in Canada's big cities and is a very real problem here in Vancouver. Just because crime figures (dodgy figures at best) have gone down does not mean we become soft on crime. If it were up to me anyone convicted of a gun crime, child abuse or sexual assault would be jailed for life and drug dealers and the likes of Clifford Olsen and Willie Picton would be hanged.

    For far too long the victims are the ones who actually suffer and its about time the soft approach was toughed up a little.

    BTW Eric. Thanks again for making my day more interesting as your blog is the first thing I read each morning. :)

    ReplyDelete
  20. Looks like someone has fallen hard for Con propaganda. I live in Toronto, and walk to work through an area that used to be unsafe. I have never seen or heard a gunshot, and feel safer now then I have in 20 years.
    It seems to me that the only people worried about crime in big cities, are people who live in distant, rural areas

    ReplyDelete
  21. People don't seem to know their crime statistics.

    Starting in the 1960's the violent crime rate increased dramatically. Around 1990 it flat lined and even trended downward slightly.

    But there are still 300% more violent crimes today than there was in 1960.


    So its like going from 2 murders a year up to 10 murders.

    And then celebrating and giving up the fight on crime because it dropped from 10 down to 8.

    ReplyDelete
  22. Shadow said...
    People don't seem to know their crime statistics.

    Starting in the 1960's the violent crime rate increased dramatically. Around 1990 it flat lined and even trended downward slightly.

    But there are still 300% more violent crimes today than there was in 1960.


    So its like going from 2 murders a year up to 10 murders.

    And then celebrating and giving up the fight on crime because it dropped from 10 down to 8.

    08 APRIL, 2011 16:11

    Well said!!!

    ReplyDelete
  23. wow! NDP is stronger in Quebec than in Ontario! How distorted the representation!

    ReplyDelete
  24. Hi Éric,

    Thank you for your awesome work on this website! I'd like to suggest an additional column in your riding projections to show the spread between the two leading parties/candidates in each riding. I think it'd be relatively easy to implement -- in Excel, you can use MAX([range])-LARGE([range],2) to get the difference between the highest and second-highest numbers in a given range. And it would make it easier to see at a glance all the key ridings that'll determine the outcome of this election.

    Once again, thank you for your awesome work!

    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.