Here’s a poll that will delight and dismay our readers, c’est selon.



  1. Ontario wanting to vote for the federal liberal party is not news, its a historical fact.

    For all the people whining about Alberta voting conservative as a block, at least Albertans have elected liberals in the last decade. The same can’t be said about Toronto electing conservatives. In fact, Mayor David Miller should run for leader of the national liberal party of Toronto.

  2. Looks like the poll isn’t available yet on the Strategic Counsel site, but that’s a pretty amazing drop.

    Here are the 20 Ontario battleground ridings they’re polling:

    Parry Sound-Muskoka
    St. Catharines
    Hamilton East-StoneyCreek
    Thunder Bay-Superior North
    Thunder Bay-Rainy River
    London West
    Mississauga South

  3. Nanos, and Harris Decima both have Harper edging up again after a small dip (the dip is caused by weekend polling which undercounts CPC voters).

    Watch Nanos continue to go up. Mondays will generally be a bit of reprieve for Liberals as the trailing weekend polling incorporates the Friday night, Saturday and Sunday. By Thursday the party’s over. Expect this rough trend to continue. Look to the week over week trends for the real picture.

    Any single poll with a drastic swing over a short period is almost by definition, an unreliable outlier.

    Focusing on patently unreliable outliers are…well, unreliable, and also underscores a certain level of desperation.

  4. kody: “the [Nanos/Decmia] dip is caused by weekend polling which undercounts CPC voters.”

    Why is that?

  5. Because you’re desperate, Jack.

  6. *Why is that?*

    They’re all out hunting moose? ;-)

    Oops – wrong election…..

  7. Jack,

    the same (rough) phenomenon can be observed in US tracking polls.

    Generally (again we’re talking statistics not absolutes) younger folks tend to be more left, and older more conservative. Families with children tend to lean more conservative, than singles. Again, the difference doesn’t have to be that significant to move a poll a couple or three points.

    Weekends are busy for families. Taking the kids to hockey tournaments/ shopping ect. If you have kids you know there are no “weekends off”.

    Also, rural folk (far more conservative) often go into town for supplies/groceries on weekends.

    Therefore they respond less and are undercounted.

  8. Paul,

    I think that snark was unwarranted. Jack asked a question which on its face appeared based on a genuine desire to know the basis of my proposition.

    I gave him my answer.

    A little touchy on the “desperation” comment, which is obvious. Focusing on outliers means one wishes to look beyond that which is probable, hence eager for the improbable.

    Are you eager for the improbable, Paul? Is that the basis for your snark?

  9. I just don’t believe in “math”.

  10. Aha! Well, if it really is an empirical phenomenon I don’t want to write it off. But I don’t see why single people necessarily sit around smoking dope all weekend waiting for the phone to ring, or why (if rural) they too, leftists though they may be, don’t have to go get groceries.

    Furthermore, if true, it might explain province-wide tracking polls, but it can’t explain Battleground Riding tracking polls, right? Because there the demographic is constant.

  11. Alas!

    Still, as long as people are determined to elect Green and Orange folks, there’ll be enough Blue…

  12. On the weekend dip thing, if true, I wonder if people don’t just feel better about life on the weekends and thus less inclined to vote for the “this’ll show ’em” party. Just wonderin’.

  13. Jack,

    could be. There’s likely all kinds of spurious and direct correlations intermingled. Again, I also don’t wish to overstate it. These are just weekly dips and rebounds.

    However, if I had to put my own money on the line I’d look for the Thursdays to be my benchmark.

  14. While it’s a common claim that weekend polls favor the Democrats, there isn’t much hard evidence to support that idea. One of the best studies of this question was conducted by two polling experts at ABC News. Gary Langer and Daniel Merkle looked at the data from ABC’s tracking polls for the last three presidential elections. They compared results from people reached on Sunday through Thursday with those reached on Friday and Saturday and found no difference. Among the Sunday-to-Thursday people polled in 2004, 49 percent supported Bush and 46 percent supported Kerry. Polls of the stay-at-home, Friday-to-Saturday crowd produced similar numbers—48 and 46.

    Saturday-Night Democrats
    Are weekend polling numbers really skewed?
    By Daniel Engber
    Updated Thursday, Nov. 2, 2006, at 6:16 PM ET

  15. Aha!

    I must say, I do wish the government would just outlaw polling. It adds this weird postmodern meta vibe to both news and politics. Would be much healthier if every reported NDP/Green statement weren’t followed by a “of course, they have no chance to form the government” caveat, if Dion weren’t quizzed about his unpopularity daily (how cyclic is that?), and if the Tories didn’t have to run against public fear of their own victory. Of course it would put all these honest pollsters out of a job, which nobody wants to do, but I do think it would clear the air. Good excuse to use the Notwithstanding Clause, too.

    Anyway (he said without missing a beat) the big thing is this 14% swing in the Ontario battlegrounds . . .

  16. Wow…yet another poll. Bestill my heart.

    Does this mean bad news for Tony Clement, or is just Mike Duffy’s main talking point for tomorrow? (I love Mikey, but does every news anchor really need to gush over their organization’s polls as God’s gift, whereas the others are all crap?)

    I learned a few elections ago to take all polls in aggregate; including tracking polls, before drawing any conclusions. In this case, as the polling isn’t national, or even regional, but riding specific; it really doesn’t seem to offer any overarching narrative.

    But that’s just me.

    In five days, if the aggregate of the national polling (weekends included) points to a neck-and-neck horserace between Dion and Harper; I’ll give credit where credit is due. If not, peraps we should all treat this like Jeff Rubin’s prediction of oil hitiing $200/bbl. A big prediction, best forgotten.

  17. Everybody take a chill pill. I think all you can really say at this stage is that the Cons are ahead of the Libs by some (highly) uncertain amount.

    I always find it amusing that ALL pollsters (I admit being a fan of Nanos) don’t plot their error bars when they compare polling results over some time period. This is standard procedure for any half-decent scientist/statistician.

    They probably avoid this because, you know, it demonstrates how uncertain polling really is.

  18. Looks like kody just wet his pants…

    Spin spin away!


  19. Austin,

    to the contrary. I’m aware of that study and it’s simply not a good one.

    I’d go into it further but I suspect I’d be preaching to closed minds.

    In the meantime I suggest you take the advice of others here (who you don’t presume to be a lying partisan con “troll”) and not put stock in this single outlier.

  20. The battleground approach — though simple and I hope cheap — is kinda irrelevant. You can’t put any significant weight in any one poll, but one that only measures 20 ridings? (And if the Tories merely won each of those ridings by a mere five points, how unhappy should they be?) What I would welcome is a poll that throws out 70 or so polls — the nation outside of Toronto (the 416), Montreal and Alberta. We know that the Liberals and the Tories both have about 30 seats guaranteed going in; now tell us what everyone else is thinking.

  21. I’d go into it further but I suspect I’d be preaching to closed minds.

    Oh, Gawd. Can DaKody’s parents please cut off his Internet already?

  22. I’m aware of that study and it’s simply not a good one.

    I’d go into it further but I suspect I’d be preaching to closed minds.

    Yeah, it’s biased. ABC News, as anyone east of Newfoundland and Labrador knows stands for Anything But Conservative News.

    Liberal media!!!

  23. Puhahaha~!

    Funny how things that contradict the narrative, is an “outlier”.

    Just remember to wipe up after you finish, k?


  24. Garnet: “The battleground approach … is kinda irrelevant. You can’t put any significant weight in any one poll, but one that only measures 20 ridings?”

    They’re doing BC Battlegrounds, QC Battlegrounds, and Ontario Battlegrounds, so it obviously each only has relevance in those provinces.

    But I’d say it’s more significant, not less, than a province-wide poll, n’est-ce pas? In terms of predicting seats, I mean. As it is, poll info from, say, Ottawa-Vanier isn’t of much interest, because it’s guaranteed the Liberals will win there; ditto whatever’s a safe Tory or NDP seat. What matters is the “marginals,” as our British comrades say. The boxing rings. The battlegrounds.

    I think my riding (Trinity-Spadina in TO) is a battleground, actually, and we’re not on the list. (Held by Olivia Chow but with a 3700-vote margin in 2006.)

    Gotta love the term “battleground,” to boot. Reminds of Darius III leveling the battlefield of Gaugamela.

  25. Kody says:

    “Also, rural folk (far more conservative) often go into town for supplies/groceries on weekends.”

    I am rural. It doesn’t take me the whole weekend to go to town for ‘supplies’ (sooo… pioneer-like)…. then again I am not conservative either.

  26. Bipoller disorder (not to be confused with the more treatable bipolar disorder) is a mood disorder that causes radical emotional changes and mood swings, from manic highs to depressive lows, in reaction to the latest poll results. The majority of bipoller individuals experience alternating episodes of mania and depression, depending on how they perceive the most recent polling numbers.

    In severe cases, the bipoller patient will become delusional, espousing the irrational belief that he can extrapolate certain (often detailed) conclusions from the polling data that clearly do not exist. At this juncture, arguing with the patient will only cause him to become defensive and agitated. Instead, the treating physician should gently remind the patient that yet another poll will be out soon, and it may or may not conform to his current beliefs.

    While treatment options are limited to symptomatic relief, the long term prognosis for for those afflicted with bipoller disorder is excellent. Symptoms nearly always abate on or shortly after Election Day.

  27. Incredible. These Ontario swing riding voters say it’s important to give smaller parties like the NDP and the Greens a voice. To back that up 33% of them are voting Liberal. Again.

  28. Raging Ranter:

    Can I get treatment now? Election day is too far away.

  29. Some polls are good news. Some are bad news. This one is totally meaningless.

  30. Well, I’m glad to learn about the weekend dip for Harper. He called the election for the end of a long weekend, so we can hope that one will be a long dip.

  31. I keep going back and forth on the Globe approach. On the one hand, in a country liee Canada using Single Member Pluralities, nationwide polls do not mean a heck of a lot (or else why is the Bloc going to win 30-odd seats off 8% of the vote and the Greens stand to win none of 9%). So looking at those ridings that are close seems like a better idea – I think the problem I have is that they are looking at ridings that are close in a three way matrix.

    So we have close Con-Lib, Con-NDP, Lib-NDP – if we only had one set of pairs, then the poll would be very useful, but given that in many of these close races, the other parties get very few votes, the poll as a whole cannot tell us who is winning (take a look at Vancouver Island North, for example, which splits somewhere around 41% NDP, 40% Con, 11% Liberal).

    I would love to see someone put up the money to do a province by province poll with a large sample in each province (500 – 1000, depending on size)

  32. Chris B, the good news is that Elections Canada will be doing a truly immense poll in mid-October. I’m told their impartiality is in question, though.

  33. What if the Jews vote first (owing to Election Day being on Sukkot) and then Harper pulls the rug out from under us and goes Hamas? That would be tricky.

  34. I heard someone from Strategic Council on the radio the other day, pointing out what one of their polls meant. They did it again in the Globe article “the Tories’ decline in Ontario reflects the fact that all parties are attacking them in advertising and campaign stories.”

    Does the fact that you compile the statistics somehow give you a better insight into the minds of the people polled? Because when those people voted Harper as the ‘nastiest’ candidate, what they really meant was he was the toughest. (I’m not sure if the poll question was who is nastiest or who is toughest, but apparently, according to Strategic Council, they mean the same thing.)

    What?? My answer to who is nastiest is completely different from my answer to who is toughest!

    Did they actually ask the question, “have you changed your previous opinion regarding the Conservatives because all four of the other parties are attacking them in advertising and campaign stories?” Because otherwise, how can they know an individual’s reason?

  35. Well as long as Elections Canada conducts it’s polling during weekdays when parents, those most renouned Conservative voters, aren’t overly busy. What with the kids in school and doing homework and piano lessons and dance lessons and their own jobs it’s like a bloody holiday in the South Pacific all week. Thus they’ll be able to produce an accurate, completely unbiased poll. Balance.

  36. Jenn,

    I agree fully with you. It drives me nuts when pollsters think they can read the minds of respondents, or that they possess some super-human insight into the behaviour of voters – when they have little more to work with than the same numbers we do.

  37. Can I get treatment now? Election day is too far away.

    There are numerous over-the-counter remedies available at your local liquor store. These products are so effective that most patients choose to continue with treatment long after the election is over.

  38. Frankly I think normal people do Not answer poll questions on the phone .. period.
    All poll Numbers from Fantazyland .

Sign in to comment.