Friday, December 30, 2005

Turning of the Tide

SES is the first polling company to have some results from what has been a truly horrific week for the Liberals between Klander and Income Trusts.

Today's results:

Lib 35%
CPC 34%
NDP 15%
BQ 13%

In itself, these numbers aren't the end of the world. But, consider a few things:

1. SES has consistently shown the Liberals well in front throughout this campaign.

2. Yesterday, SES had the Liberals up 38-32. If we assume all three days in the rolling poll were roughly equal, that means that the single day totals for Thursday (albeit with a high margin of error) were probably in the neighborhood of 29% for the Liberals and 38% for the Tories.

3. I can't remember the last time I've seen the Conservatives at 34% in the polls. If they get 34% on election day, they'll form government.

20 Comments:

  • These kinds of swings don't mean anything unless they are permanent. On Gomery, there was a ONE day swing to the Conservatives that quickly went back to the Liberals.

    Making them permanent is a matter of Stephen Harper exciting people with hope and possibilities now he's got their attention.

    I won't put it past him, but it's not a slam dunk either.

    By Anonymous Anonymous, at 12:58 PM  

  • Definitely. We've seen the knee jerk anti-Liberal spikes before and they always fade away. If all the polls by week's end confirm the gap is still only a couple of points, there's still lots of time for Paul to turn things around.

    By Blogger calgarygrit, at 1:04 PM  

  • This is not really a one day thing. The Conservatives have been very slowly gaining. Most likely todays poll result was bit of an overshoot but you can see a trend from the beginning of the campaign. Just ask yourself by comparing this campaign with 2004, just how many Conservative will switch their vote to Liberal or vice versa. I am sure that not as many voters are as scared this time. One more thing not measured well in these polls : given how to campaign is going, how many Liberal or Conservative will actually bother to vote. Someone might not want to vote Conservative but he might not want to get of the warm couch on Jan 23 either to vote for these dimwits.

    By Anonymous Rob, at 1:20 PM  

  • I hate polls, they're so deceiving. Certain partisan newspapers like to "interpret" the polls for their readers too. They spin the results, it's bad news. They should present the poll data in raw numerical form and let the people interpret it on their own. Instead they become "gatekeepers", and it ain't right.

    By Blogger Charles J, at 1:32 PM  

  • Rob, if you look at the full trend lines, the Liberals shot up immediately during the campaign, likely on "homosexual sex marriage" (to quote Rob Anders). All that has happened since is a gradual shift back to the pre-writ numbers, which means basically so far nothing has happened.

    Right now the numbers are a one day thing because it's only been one day.

    Granted now that people are slowly waking up to the election, having a huge scandal of actual governance (*) is not likely to be good for the Liberals in real terms.

    (*) As retarded as "beer and popcorn" and "chow chow" were, those kinds of mistakes don't really matter. People focus on actions since words are only worth the stale air they float on. A major fuck up like this is a real matter of confidence. --anon 10:58

    By Anonymous Anonymous, at 1:39 PM  

  • I'm skeptical that the "tide has turned". My view is that Ontario voters are like fields of wheat. If you blow on them really hard in one direction, they all kind of lean over that way, and let you think you've made a permanent change in their orientation.

    But the minute you stop blowing, they spring right back up to where they were before, i.e., ready aye ready to send another 75-100 Liberals to Ottawa.

    So my suspicion is that this isn't a turning of the tide so much as one of those stiff breezes that occasionally hits southern ontario. We'll see if it strengthens to a gale. I think not.

    By Anonymous andrew, at 1:48 PM  

  • To the extent that one can put faith in the SES numbers, I put the Canada and Ontario Liberal and CPC numbers into Excel, and put on the trendlines.

    Neither looks good for the Libs, but the Ontario numbers must truly be frightening for them. Equal share crossover at Jan 20th, if current trends hold up. That's NOT good news.

    Even the Maritimes shows a crossover well before that.

    Not on my website yet. Haven't figured how to get the Excel charts onto Linux and WordPress.

    By Anonymous Erik Sorenson, at 1:52 PM  

  • Actually, Ipsos Reid had a poll done on the 28th which also showed the Libs at 33%, the Cons at 32% with the polling trend towards the Tories. In Ontario, the two parties were statistically tied, and the SES poll today shows pretty much the same trend. Where it does differ is in the Atlantic provinces, which shows that the two parties are tied at 40%.

    So... using the Hill and Knowlten predictor at:

    http://predictor.hillandknowlton.ca/default.asp

    The landscape would look like the Conservatives with 121 seats, the Libs with 114, the NDP with 13, and the Bloc with 59.

    The Ipsos poll show's a significantly different result with the Libs at 101, Cons at 117, NDP at 26, and the Bloc at 63, with a seat for the Greens.

    Fun stuff eh?

    By Blogger Joe Calgary, at 1:55 PM  

  • Sorry, the Ipsos poll was completed on the 24th of December. I don't place a huge amount of faith in the pollsters either, but it's darn hard not to look at them.

    By Blogger Joe Calgary, at 1:57 PM  

  • Erik, select the table / chart in Excel. Copy it.

    Run Paint; it's under Programs / Accessories. If you want to run it from the command line, it's called pbrush.

    Set the image size to 1x1 pixels. Go to the Image menu and run the Attributes command. Change the width and height to 1 each.

    Paste the image.

    File, Save As the filename. Make sure you set the file type to PNG in that dropdown list.

    Then you can post the table in your blog.

    By Anonymous Anonymous, at 1:58 PM  

  • Just a note, I am pretty sure that this is actually the first poll out since before the break. Its data is from 23, 28, 29. The last poll was I believe from before the break and was 20,21,22.

    By Blogger Political Nobody, at 2:21 PM  

  • I don't put much faith in SES polling numbers. They ask responders to rank their choices rather than pick one or say undecided. This has the effect of exagerating the lead of whoever has momentum. I still think that the Grits and Tories are basically tied, and don't forget all these changes have been within the margin of error, not to meniton that whenever they publish data from regions the sample size is far too small (less than 500) to take seriously.

    By Anonymous Andrew, at 4:05 PM  

  • Could it be that the WHEELS OF CORRUPTION are falling off the Liberal Bandwagon . . . one can only hope.
    Did you know . . . Paul Martin loves Canada so much that he keeps his millions in Bermuda, and has his family business flagged in a foreign port . . . wants to keep it all away from Liberal Corruption I guess . . .

    By Blogger EX-NDIP, at 4:29 PM  

  • SES doesn't mean anything seat-wise until its rolling polls add up to a shift in South-Western Ontario and Southern BC (and Nova Scotia, to a lesser extent). I can't see that just yet; check out the margin of error in their recent 400 person survey (i.e. 6-10% by region).

    By Blogger matt, at 5:16 PM  

  • Again Andrew blames Ontario. There's as much Conservative voters in Ontario (if not more) as there are in ALL of ALBERTA. It's the fault of the electoral system it is not reflected in number of seats.

    Ontario seems to have no problem with John Tory (he's up 50% in polls now), Mike Harris (he was elected to two majority governments over Pinocchio), or Peter MacKay or Belinda (when she was a Tory) and so on.

    By Blogger mezba, at 9:42 PM  

  • Grit,

    Interesting poll. The undecided numbers seem to be slowly dropping, but they're still well above 15%. Too many undecideds to say anything for sure yet.

    By Blogger Andrew, at 9:47 PM  

  • No longer a Harper menorah-ity;
    Now it's a CPC majority !!!

    N.B. Ya red it heah foist, etc.

    What with the latest Ipsos-Reid survey revealing 37% of voters volatile and prone to switch, and the SES/CPAC recent overnights showing leaps-and-bounds gains for Harper vs. Junior--both in terms of beauty contest attitudinals re: PMabilities and pop vote intentions--the Molson Duvernay computer projectionware algorithms have churned out a Tory majority gov't come 11:10 pm EST, Jan. 23, 2006.

    Having FIRST suggested publicly via Azerb's comments section --ie. well ahead of WK, who, not unlike Sheila Copps in her former riding, has been whispering strategic and tactical advice to Team Harper--that l'il Stevie was headin towards a menorah-ity PMship (see recent G&M photo of former reform Prestone henchperson Harper replete with stapled on beany lighting a Chanukah candle), I now believe that the near-perfect campaign message execution and superbly cynical aggregate vote bidding efforts by Harper will pay off big time with a majority win.

    Without boring you all with any further specific poll analyses (see comment # 13 chez http://thestar.blogs.com/azerb/2005/12/although_i_am_a.html#comments ) let's cut to the finish line in this year's reptile run:

    ...in the context of a first-half wherein Harper and Martin have gradually switched their numbers over the course of the first month following the first debates--which were held at a time when a plurality of voters in ROC found neither of the two putative majority party leaders leaders to be particularly credible, persuasive, or impressive--a likely plurality of voters will tend to tally up the cash-forthcoming promises and vote according to their own personal pocketbook perceptions...

    If I am correct in this surmise, then Harper wins hand down due to the following 3 'keepin'it simple' promises:

    1.) two cents worth off the GST--one upfront, the other over time, a weasly promise which will be computed in all manner of households from the Rock to Tofino and up to Yellowknife;

    2.)$1200 per family in direct payoffs per kid under 6, which honours Harper's Bushite base of stay-at-home mom families and evangelicals and other ethnic and religious-conservative voting groups as well as the gullible working poor and esp. those parents--incl. many rural households--who don't currently have access to or want access to publicly funded day care places (these folks will ignore the reality that the enactment of such a promise will result in the callous rescinding of the already announced agreements by the Feds handing over huge bucks to fund/create additional public daycare places;

    3.) ...and the most ideologically neutral and brilliantly conceived/ wedged campaign giveaway goodie to date, which will cut across all party preferences and draw lower-to-middle and even upper-middle income household members--gen Y thru seniors' demos--to the tories:

    Harper's 'green-friendly' promise to subsidise, via tax deductions, up to 16% of personal annual expenditures on monthly mass transit passes (thus excluding many students who have declared themselves financially independent of their parents plus many, many others who don't pay taxes or who survive on social benefits and/or poverty level fixed incomes.

    The latter promise will serve to sway significant swatches of switch voters (who now total over one-third of the electorate) in a huge array of suburban and satellite exurban centres coast-to-coast !

    By Anonymous MoDuv, at 11:50 PM  

  • Margin of error means that no poll on its own -- or even two polls -- should be taken very seriously. If I'm doing the math right, a 3 percent margin of error means that a true Liberal gain in relation to the Tories of as much as 6 percent over a given period could be reported by pollsters as a LOSS of up to 6 percent, even if the polls involved are within the margin of error and do not include a 1-in-20 aberration. If the Liberals are reported as being at 35%, they could be as low as 32% or as high as 38%. Tories at 34% means anywhere from 31% to 37%. So the Grits may have a 7% lead or be behind by 5%.

    When we look at a breakdown by province or even region, the larger margins of error mean that the party figures are quite useless. Then we have to remember that party support is very uneven across the country and even from riding to riding. I'm in Victoria, which looks to be a two-way race between the NDP and the Liberals. Next door to the north is Saanich-Gulf Islands, probably a good bet for the Tories, but nonetheless a four-way slugfest in which the Green candidate got more than 10,000 votes in 2004. Overall party percentages for the province mean less than nothing in judging who's going to win these ridings.

    What we can say with confidence is that, given the many close riding contests, small changes in party support could result in a significantly different political landscape after January 23, although a minority government is still a likelihood.

    By Blogger Aeolus, at 12:42 AM  

  • Aeolus,

    Good analysis. We might also see a repeat of the last election, where Harper is rising and rising and then Martin appeals to all NDP, Liberals to hold off Harper, and then resulting in a minority Liberal, weak NDP, stronger CPC. Result, Liberals boot Martin, elect a younger, tougher and popular leader, another election in another year and Liberals majority.

    We can dream.

    By Blogger mezba, at 11:42 AM  

  • Congratulations Friend for your excellent blog on make money!Keep up the good work!
    If you have a moment, please visit my site:
    make money
    I send you my warm regards and wish you continued success.
    Have a nice day! :-)

    By Blogger James Baker, at 1:09 AM  

Post a Comment

Links to this post:

Create a Link

<< Home