06.11.2014 04:35 PM

Ipsos just now: Ontario Liberals behind in every region

…with those most likely to vote. Holy shit la merde!

They’ve got a huge sample – two thousand people. And look at what’s happened to the NDP vote! If this is a trend, and not a blip, who is going to be leader of the Opposition?

And, continuing with the breathless questions, what does it all mean, Virginia? It means GOTV tomorrow determines who wins. Full stop.

Ipsos, with my emphasis added:

“Examining the regional vote by those most likely to vote:

· In the 905, the PCs (43%) pull ahead of the Liberals (30%) and NDP (23%).

· In the 416, the NDP (38%) are in a statistical tie with the Liberals (35%),while the Tories (24%) trail.

· In Southwestern Ontario, the NDP (39%) are also ahead of the PCs (32%) and Liberals (25%).

· In Eastern Ontario, the Tories (43%) do well against the Liberals (29%) and NDP (24%).

· In Central Ontario (based on a small sample size), the PCs (45%) pull ahead of the Liberals (35%) and NDP (17%).

· In Northern Ontario, the results (based on a small sample size), the PCs (37%), Liberals (34%) and NDP (28%) remain competitive.”



  1. Ty says:

    “In Southwestern Ontario, the NDP (39%) are also ahead of the PCs (32%)[…] In Northern Ontario…the PCs (37%), Liberals (34%) and NDP (28%) remain competitive.”


  2. Matt says:

    Don’t you think Wynne’s constant “NDP supporters have to vote Liberal because only the Liberals can stop Tim Hudak” over the last few days just reeks of arrogance?

    All she’s done, and the Ipsos numbers seem to show this, is sufficiently piss off NDP voters to get them motivated to go vote NDP.

  3. sezme says:

    Well, that’s pretty much the craziest polling result on the eve of an election I’ve seen. Especially forgetting their “likely to vote” model, this:
    Liberals (33%), PCs (31%), NDP (30%)
    is a pretty absurd result.

    I’ll take it with a grain of salt until tomorrow.

    • Matt says:

      Some, or maybe even all pollsters are going to have egg on their face Friday morning.

      • Al in Cranbrook says:


        Frankly, I think a growing number of people simply don’t want to engage pollsters any more. One stat we never hear is how many calls they have to make to get one interview. I’ve heard in the past that it’s a remarkably high number.

        I’ve also heard that, historically, Conservatives tend to be the least likely to want to share their intentions with pollsters. I can believe that, and I’d also believe the opposite is likely true of left leaning voters, a lot of whom are younger, and thus tend to look for validation wherever they can find it.

        • Kaspar Juul says:

          I think, I’ve heard, I’d also believe. So much for facts. Just projecting opinion again.

          Post less Al. Make them actually worth something

          • Lance says:

            So he’s not allowed to post his opinion? Why, because it rubs you the wrong way? His opinion has to jive with your’s to be “worth something”, therefore he should “post less”?

            A hypocritical paragon of the “tolerant left” speaks again.

          • sezme says:

            @Lance: Jibe. The word is jibe. (Sorry for the pedantry; no disrespect implied.) Also, it’s “yours”.

          • Al in Cranbrook says:

            Kaspar’s philosophy on existence…

            “I have a computer, therefore, I am.”

            Thus, to him, “I think”, is an abstract concept.

            Welcome to the 21st century, eh?

          • donna says:

            Where “less is more”? Follow your own dictum… ya dick!

        • Lance says:

          Thanks for the grammar and spelling reminder, Sezme. I knew that of course, however, since I’m not writing for my employer here or composing a dissertation (thank goodness, LOL), I tend to slide a bit at times.

  4. paulsstuff says:

    I can’t help but feel I’ve seen this movie before. The attack ads, the desparation, the shrill voice asking for supporters of the other party to support them instead to stop the spawn of Satan winning the election.

    Oh wait. I’ve got it. It’s Paul Martin’s 2006 federal campaign playing out in an Ontario election, with the role of Paul Martin now being played by Kathleen Wynne. Who doesn’t love a great comedy? Call it Dumb and Dumber 2.

  5. AC says:

    Abacus (also an online poll): OLP 36%, PC 36%, NDP 23% (likely voters).

  6. Coelocanth_Jones says:

    Having taken a little time to read it over, the conservative lead up north, with NDP in third, seems to to bee the oddest result here. In any case, if something like this is brewing for tomorrow, I hope Wynne’s people are hugging themselves particularly hard

  7. Corey says:

    A few issues here…
    1. This is an online poll and those are tricky. Ipsos has reported higher NDP numbers than the other pollsters for a while now.
    2. Every other pollster has NDP support dipping over the same period that Ipsos has them surging. Who’s right?
    3. NDP GOTV is not the best in Ontario… especially without their usual union backers.

    I guess we’ll see tomorrow night… what a polling mess this election has been!

    • Doodle says:

      The Ipsos likely NDP numbers climbing since the debate are consistent with what Abacus and Angus are showing, so it’s certainly plausible result.

  8. Al in Cranbrook says:

    I give up. All bets are off.

  9. Lance says:

    In Nickel Belt I have seen NOT ONE PC sign ANYWHERE. NDP and the LPC fight it out over the strong labour vote in Sudbury with the PC barely even a blip. I know there are other Northern ridings, but the PCs ahead?In I’m not sure I buy that.

  10. Chris Stockwell says:

    I believe you saw this coming Warren. You saw how Wynne was acting the last week of the campaign. Something was very very wrong in Liberal land.

  11. Corey says:

    Now another poll from Léger has OLP 37, PC 37, NDP 20…. why is only IPSOS showing NDP gaining while all the others have NDP dropping??? This has to be the most confusing election in terms of polls we’ve ever had.

    • Al in Cranbrook says:

      Leger poll, PDF…


      In the age groups from 35 up PCs have biggest lead. These usually are the ones most likely to vote.

      A lot of both Lib and NDP support is among the younger demographic.

    • Doodle says:

      That poll shows some very interesting numbers among Liberals who are not fully committed and what they would vote for if they changed their mind. And it kind of falls into place with that whole NDP gaining thing that Ipsos and – to a lesser extent – Abacus is showing.

  12. Coelocanth_Jones says:

    I find it hard to believe that the PCs are ahead up north, with the NDP in third. Otherwise, if anything similar to this unfolds tomorrow, I hope Wynne’s people are hugging themselves extra hard

    • Coelocanth_Jones says:

      Forgive my repost, warren, I had skipped over my original comment and assumed that something went screwy in the captcha

  13. Todd Robson says:

    Unless pigs are flying over the Big Nickel … there is no way that Northern Ontario sample is correct.

    • Matt says:

      Well, they did say it was a smaller sample size.

      But, it might be accurate. It would explain the Liberals and NDP suddenly ragging on Hudak for not making a single campaign stop in the North.

      I thought it a bit wierd they would wait until 1 day before E-Day to start that line of attack.

      • Lance says:

        I’m almost positive that is why Hudak didn’t even bother. “Prior commitments” so he cannot attend a Northern debate? C’mon. He knows the North is dead to him, so these polling numbers about the Tories being ahead in the North make absolutely NO sense.

        Is signage any indication? Maybe, maybe not; but when you see NO PC signs in one riding, and nothing but a sea of orange with a scant tinge of red, well, you don’t need to stretch to draw the conclusion.

    • Lance says:


  14. Bill From Willowdale says:

    Looks like a wild ride tomorrow. It’s going to be an exciting night for political junkies like us.

  15. Dave Ruddell says:

    I’ve been getting telephone polled (not live) a lot. I think four times in the last two weeks, with the question being which local candidate I’d choose (this is in Beaches – East York). The follow up question then asks, as a second choice, which party leader I’d pick, rather than the local candidate. I don’t really understand the methodology here. The strange thing is, after answering the second question (press 1 for…), there’s nothing; no thank you, no goodbye, nothing.

    The opening part of the call identifies the polling firm and contact info, but it goes by so quickly that I don’t have a chance to write it down. What I do know is that I don’t recognize the name of the firm. Would I be correct in assuming that this would mean it’s an internal poll by one of the parties? If that is the case, I can tell you that it’s not the Tories, since they have no chance. That would mean it’s either the Liberals, because they think they have a chance of winning, or the NDP thinking they have a chance of losing what I had thought was a safe seat. If the OLP can pick up a seat like Beaches – East York, then I think it will be a very good night for Wynne.

    • Matt says:

      I’m in Scarborough Centre.

      I’ve had 1 call from the PC’s, a recorded message. None from the Liberals, NDP or Greens.

      And what’s wierder, I got only one piece of election advertising in the mail which was again from the PC’s on Monday.

    • sezme says:

      I got the same call as you, Dave. Not sure who it was.

  16. Clive says:

    Meanwhile Ekos predict a Wynne majority. Someone’s going to have ‘splaining to do tomorrow.

    • Emil E says:

      Me too… and it’s based on an energized 1.2 million public service union voters flocking to Wynne Liberals plus Dippers who want to Stop Hudak. I believe this is the underlying current in this election that cannot be gauged by polling. Now we wait to see if the unions do GOTV for their Liberal saviours.

  17. Clive says:

    Warren, while we don’t agree on much, given your expertise I wonder if you have any thoughts on an effect that was seen in Britain in 1992 and, I suspect, also in Ontario in the second Harris era. In 1992 in the UK, with the Tories in disarray and with a fairly weak leader, everyone expected Labour to cruise to victory and the pollsters predicted this. But it turned out that in the polling a number of people hadn’t been truthful because they were embarrassed to admit they were planning to vote for Major. He did way better than the pollsters predicted for that reason. Now, this predates online and IVR, so people being polled were talking to a live person which increases this effect. But have you ever seen anything like that in play and do you think that could play to Hudak’s advantage tomorrow?

    • sezme says:

      Why would Ontario PCs be embarrassed to admit their voting intentions? If anything, given the narrative of this campaign, I’d expect OLP voters to not wish to fess up to supporting a “criminal, corrupt” party.

      • Doodle says:

        Depends on the specific demographics. I personally know of a few public who will not openly admit they are plugging their nose and voting for the Tories for fear of reprisals from their union or co-workers who can best be described as militant. That, and there is historical evidence to believe that pollsters have a harder time gauging the true number of people who are willing to vote for the Tories, versus over-estimating Liberal and/or NDP support. Happened in the 2011 federal election, and hell it even happened most recently in the Niagara Falls byelection that was a lot closer than the polls predicted. This may be an anomaly in methodology rather than people actually expressing support. But I don’t run the polls or create the reports. I merely read them.

  18. Karl says:

    The EKOS likely voter numbers are so far away from iPCos numbers that someone (perhaps both) will have a lot of egg on their face.

    So let’s presume for the moment that they are both outliers and average all of the six polls on an unweighted basis using likely voters where stated: EKOS, L41%, PC 33%, NDP 17%; iPCos, L30%, PC36%, NDP30%; Abacus L36%, PC34%, NDP26%; Angus Reid L34%, PC36%, NDP 24%; Leger L37%, PC 37%, NDP 20%; and Forum L42%, PC 35%, NDP 19%.

    That gives us a six poll average of 37% Liberal, 35% PC and 23% NDP. Given relative voting efficiencies, that’s points to a reasonable Liberal plurality, with creditable NDP representation. And by the way – 2011 results were 37.65% Liberal, 35.45% PC and 22.74% NDP.

  19. Karl says:

    Sorry, I should note that if you were to drop both EKOS and IPSOS as outliers, you’d get essentially the same result: Lib 37.25, PC 35.5, NDP 22.25 and again, each within a few tenths of one percent of 2011 numbers.

  20. MoeL says:

    Live and IVR polls randomly call telephone numbers. Could someone explain to me how an online survey RANDOMLY selects its respondents. If the sample is not truly random, then does it really doesn’t matter how big their sample is.

  21. Matt says:

    Regarding EKOS’ numbers:

    The National Post reported they made a “change” to their methodolgy/voter models before their last poll which put the Liberals way out front in likely voters.

    Nope. Nothing fishy there.

  22. Doron says:

    I think given the nature of our system, polling firms should only poll swing ridings and make assumptions for safe seats.

    Then show a seat projection.

    Percentages don’t mean much.

  23. Bill From Willowdale says:

    Warren, are you going to have an election prediction thread?

  24. Matt says:

    Just voted here in Scarborough Centre.

    This particular polling station was a ghost town. Just me and one other person voting. On the way out I asked how the turnout was. The lady at the entrance said “about 30 so far”

  25. TrueNorthist says:

    I am largely in agreement with those here who call for a moratorium on polling during an election campaign, but I think a week or 10 days is a much more reasonable exclusion period. Polling is getting so far out of hand that it is severely impacting the process itself. We are slowly becoming just like the USA with it’s bought and paid for elections.

  26. Jacob Trouba says:

    Wildrose Majority!!!!!!!

  27. kre8tv says:

    I look forward to the talk that will start tomorrow — as with just about every vote in the last bunch of years — in which a number of folks will say “well that’s the *last* time I’m ever going to trust the accuracy of polls during a campaign.”

  28. Vince says:

    Despite the prominent failures in B.C. and Alberta, pollsters have correctly predicted many elections since 2010 (as Grenier notes today). I would still like to see a discussion which moves beyond popular vote to discuss vote efficiency. It would be hilarious if the Libs won a majority and lost the popular vote, though that is unlikely. It would have been interesting if Ipsos had given a more detailed report.
    Interesting that Breguet (with the fullest discussion) at 2closetocall has the probability of a Liberal majority greater than a PC minority. Perfect storm has to happen for PCs to win a minority: they must take 4/5 from Libs and NDP must take 4/5 from Libs with no seats taken from them. Highly unlikely. Libs will have to lose York Southwest and not win Trinity or Davenport for starters. If they can hold Glengarry or Nepean or Brant, and poach Malton, Cambridge and Perth, they’re looking good.

  29. Adam says:


    I think you’re right. EKOS saying college/university educated are “considerably more likely to vote” seems to be pulled out of thin air, as opposed to grounded in quantitative research. Seniors and those above 45 are more likely to vote as research has long shown.

    This is the problem with media pollsters; It isn’t driven by science, it is driven by a competition among POR firms and media outlets.

    I’d hate to think EKOS would adjust their data collection to get the result they want but I’ve never heard a last-minute adjustment of method to simply capture a cohort of likely voters based on an assumption … Maybe it is grounded in scientific method.

    Either way, I don’t know what to make of any of it. I just think it would be absolutely a beautiful thing if the Jugger-NOTS lose again.

  30. Patrick says:

    And CCTV is calling for a liberal majority.at 10’15 pm. Proving bullshit questions gets you bullshit answers.

Leave a Reply

Your email address will not be published. Required fields are marked *