Friday, August 14, 2015

2015 Federal Election Link Round-up, Week 2

Week 2 begins! Polling from Week 1 was notable for its absence, which is perhaps not too surprising. Polling takes a little bit of lead time, and vacations can be difficult to re-schedule on such short notice. Hopefully we'll see some more numbers this week.

Friday, August 14, 2015

- The riding projections are now up-to-date.
- The Poll Tracker has been updated, and here is my latest analysis of the numbers (including the new poll from Mainstreet Research).
- A new episode of the Pollcast was uploaded yesterday. The guest is Dimitri Pantazopoulos, a pollster that has worked with the B.C. Liberals and the federal Conservatives. He gives us a peek at how parties are doing and reading their polls.

Thursday, August 13, 2015

- The Poll Tracker's been updated with the latest Forum poll. The impact has primarily been a downturn for the Conservatives.
- I was on Power and Politics last night talking about the effect of the Duffy trial and the debate on the polls.

Wednesday, August 12, 2015

- Here's my analysis of the latest Poll Tracker update.
- Riding projections have been updated to reflect the latest Poll Tracker update.
- The Poll Tracker has been updated with the new Ipsos-Reid poll for Global News. You might notice that the poll has a very high weight. It has run up against the cap the model has on the weighting applied to any one poll, that being 67%. The Ipsos poll has hit this cap because it is, really, the first poll to have been conducted since August 2. That makes it over a week newer than the previous poll in the model (Forum's). The Nanos poll, though it ended on August 7, was primarily conducted before the writs dropped as it is a four-week rolling poll. Simply put, the model considers the Ipsos poll to be pretty much the only current information we have, so it has a high weighting. This will not happen very often going forward, as the pace of polling will undoubtedly pick up.
- For my CBC column today, I wrote about the impact of incumbents not running for re-election. Some interesting findings, drawing on the calculations used for the model as well as the riding projections.
- The latest episode of the Election Pollcast was uploaded last night. This week's guest is Paul Adams, associate professor of journalism at Carleton University and someone who has worked both as a journalist and as a pollster. A really fascinating discussion of the way the relationship between pollsters and the media has changed over the last 20-30 years.

Tuesday, August 11, 2015

- The riding projections have been updated to align with the latest Poll Tracker numbers.
- Finally, a poll! Sure, it is a four-week rolling poll with the bulk of it having been taking before the writs were dropped (and all but one day of it done before the debate), but we've gotten desperate. Here's the link to the Nanos poll. The Poll Tracker has been updated too, but the changes have been marginal.

Monday, August 10, 2015

- Time to check-in on the polling averages for July.
- The polls are in for Thursday's debate, and the results are...oddly regional. Mainstreet Research looked at what British Columbians thought, while Forum Research looked at what residents of the GTA thought. Well, residents who were at home in the middle of the day on a Friday. Sheesh. If we can draw anything from these, it is that Justin Trudeau and Elizabeth May probably did the most to help their respective causes.
- RDI Matin Week-end, au sujet des sondages politiques.
- Following the interesting discussion with Christian Bourque in last week's podcast, I looked at the challenges this long and tight campaign poses for pollsters.
- In our regional go-around, Chris Hall and I discuss Alberta on The House.
- Debate reaction on CBC Ottawa News, starting at 14:30.

42 comments:

  1. Written reviews of the debate were all over the map. If there's a consensus, it's that May did well.

    I watched the debate. Everyone did fairly badly, I thought. There were opportunities to win for everyone, and I don't think anyone seized them.

    ReplyDelete
  2. What values are there in a poll conducted from between 10 AM and 3 PM on a Friday? Isn't that inherently distorted?

    ReplyDelete
  3. I thought Mulcair and May were the best. The few memorable lines, the ones likely to be replayed as sound-bites over the next 2 months, were all theirs. Certainly Harper managed to parry all the attacks (and, as PM, inevitably got most of them) and Trudeau came across as competent and qualified (which, unlike some viewers, I never doubted) but they didn't land any real punches.

    ReplyDelete
  4. With expectations significantly lowered for JT going into the first debate, I can see why many saw him as coming out on top...he was a lot stronger and more credible than expected given all of the "just not ready" talk. But agree with Ira, nobody really rose above the pack except May - who, let's face it, is irrelevant.

    BTW Eric - I'm sure you've been asked this ad nauseum...are you changing the name of the website after the election?

    ReplyDelete
    Replies
    1. Éric answered the name question here: http://www.threehundredeight.com/2015/07/new-ipsos-poll-battelground-toronto-and.html

      Delete
    2. I could understand the belief that Trudeau did well if not for his closing remarks. They were remarkably vapid. I also didn't like his use of "No one believes you" when responding to true statements Harper made. Maybe it's true that no one believes Harper, and maybe it's true that Harper was spouting entirely useless facts, but they were still facts. I would much prefer Trudeau make reasoned arguments. I was reminded of the moment in the 2006 debate when Paul Martin was claiming that his government would lower taxes on the lowest income Canadians, and Harper responded (entirely factually) with "The lowest income Canadians don't pay any taxes." By the third time he said it he was clearly exasperated, and verged on condescending, but he was right.

      I like these debates to be about facts.

      Harper was oddly stilted in this one. You'd think he'd be better at this by now.

      Mulcair's closing remarks worked well when he was moving quickly, but he'd clearly been told to speak more slowly to give his words gravity, and that messed up his cadence and made him lose his place in his notes. Only when he realised he was running out of time did he pick up the pace, and then suddenly he was better. His campaign team is responsible for that one; they need to get to know their leader better.

      Delete
  5. It's August; only political junkies cared about the debate, and we're the least likely to be swayed by it.

    ReplyDelete
  6. This debate was pretty mild, with nobody really standing out too much. May didn't do badly, but she was talked over a lot. Probably the most significant revelation was Harper's admission that he asked the Senate to kill a bill passed by the elected House of Commons.

    Foreign policy was definitely a weak spot, with not a single leader being all that honest. For example, nobody mentioned the issue of the Turks attacking the Kurds in last week, who're supposedly our main anti-ISIS allies. Or that it has conclusively been proven by now that ISIS gets money by selling oil to Turkey. I guess nobody wanted to embarrass a NATO ally, but it's one more reason that Canada shouldn't be getting militarily involved in this mess - wouldn't it be more honest to put pressure on Turkey to cut off ISIS's funding instead? It seems that Canada's entire foreign policy lately is to be loyal to whatever the neocons in the US want, with no independent thought.

    Ukraine's another example, with not a single party leader acknowledging that the situation there is a lot more complex than the "evil-Putin-attacks-peaceful-Ukrainian-democracy-lovers" narrative, or that almost all of the civilians are continuing to be killed by "our" side (it's one reason that I'll never vote for the NDP in this election, as Mulcair is even more of a war-monger than Harper in Ukraine, and Harper's bad enough).

    ReplyDelete
  7. Hi Eric, love your work. Just curious why the poll tracker hasn't been updated in over a week. Has there been a lack of suitable polls?

    ReplyDelete
    Replies
    1. There has been a complete lack of polls! The last publicly released national poll was conducted on Aug 2.

      Delete
  8. Face it these debates aren't about "Truth" or "Reality" !! They are all about appearance and the cheap shot. As such basically useless.

    ReplyDelete
    Replies
    1. That this works, incidentally, is why I'm not super excited about getting more people to vote.

      The ones who do are already easily fooled. The ones who aren't paying attention will be even more so.

      Delete
    2. If you aren't paying attention I think it difficult to be fooled.

      Delete
    3. I disagree. If you make a decision without having paid attention, it's easier to believe the most attractive lie you're told because you're less likely to perceive the complexity of the issue, or come across contrary evidence.

      Those flat earthers on YouTube don't hold those opinions because they're paying attention to the world around them.

      Delete
    4. How is one suppose to use sound (hearing) to determine a position when one isn't paying attention, you simply will not register what is written or said and since you are not paying attention it will be difficult if not impossible to turn any facts one hears or reads into the basis of a decision.

      As for the flatearthers on Youtube. Perhaps they believe in a flat Earth because they do pay attention and willfully choose to disregards others' opinions. The point is if one does not pay attention others' opinions, facts, context, will have zero impact for the simple reason that one is not paying attention. In short it becomes a wholly personal decision. How can one be fooled when they disregard any external material?

      Delete
    5. They're not going to learn about public policy through introspection alone. Someone gave them the ideas they have.

      Perhaps I should have said no playing "close attention".

      Delete
  9. Just saying, I was polled twice yesterday about the election: once by IVR for Forum and once online by Léger, so I guess we can expect both to be released over the next few days.

    ReplyDelete
    Replies
    1. Which should lead everyone to be cautious about how representative the sample is of the general Canadian population. 33 million people and you were polled twice in one day by two different firms???? This cannot possibly a random sample. These numbers reflect people who answer polls on any given day, based on lists that were not randomly gathered, not the general population. I'm surprised Main Street, run by Ignatief's head organizer didn't poll you as well. In a truly random sample you might be polled once in your lifetime if you were lucky.

      Delete
    2. And that is not the first time I have heard of one person being polled multiple times during an election. Some polls are even set up that way as a panel, and many polling firms use lists that are not the phone book. Just because a polling sample is adjusted to reflect census data by age, sex and region does not make it a random sample representative of the general population.

      I would be interested in having Eric ask his commenters how many times they have been polled about an election. Keep in mind 33 million Canadians, lets say 25 million can vote. If Eric gets, yeah I've been polled several times, and I think he will, we will have a greater understanding of the weakness of various polls. Its all about their lists not necessarily the methodology aka IVR versus live interview, or cell phone versus landline. It's where did the list come from, sometimes believe it or not, the list comes from a political party.

      Delete
    3. Léger uses an online panel, so choosing to participate in one of their polls is not the same as being randomly dialed on the telephone.

      I've been polled twice in my life, at different phone numbers.

      Delete
    4. I used to a member of Angus Reid's online panel, and I have been polled by EKOS on more than one occasion, but only when I lived in Vancouver.

      Delete
    5. And you have been a member of a political party have you not Eric? It also may be a case of throw enough gunk at a wall and some of it will stick. If polling companies are making lets say 100 calls to get one response, then it is also who is choosing to answer. Aka the general population does not answer polls. What we are seeing is who from the list any given company is using is choosing to answer the poll, even if it was a random computer dial of all phone numbers in Canada, and I don't think that is how many of them do it. It's too much work, and its too expensive. That's why you get the Liberals in BC having far superior polling numbers, and knowing they were winning, when the prediction was an NDP blowout. People wonder why the Conservatives are still targeting Mr. Trudeau not Mr. Mulcair. I will put forward that that is exactly what the superior polling numbers that cost lots of money are telling them to do.

      Delete
  10. It seems the debate audience was so small (just us junkies watching) that no votes changed hands.

    Mind you if somebody gaffed, we'd be seeing the clip several times a day.

    As for gaffes, I suspect McQuaig has been admonished not to indulge in scholarly speculation. A certain NDP candidate for BC PM blew a lead doing that.

    ReplyDelete
  11. "with the bulk of it having been taking before the writs were dropped (and all but one day of it done after the debate)"

    I may simply be confused, but shouldn't that read : "and all but one day of it done before the debate" ?

    ReplyDelete
  12. Thierry Soucie could you give us a seat break down based on the Nanos numbers alone? TIA!

    ReplyDelete
    Replies
    1. Sorry for the delay. So here goes:

      138 CPC
      114 NDP
      83 LPC
      2 BQ
      1 GPC

      By region, it is:

      Atlantic
      21 LPC
      7 NDP
      4 CPC

      Québec
      55 NDP
      15 LPC
      6 CPC
      2 BQ

      Ontario
      61 CPC
      35 LPC
      25 NDP

      Prairies
      23 CPC
      3 LPC
      2 NDP

      Alberta
      31 CPC
      2 NDP
      1 LPC

      British Columbia
      22 NDP
      12 CPC
      7 LPC
      1 GPC

      Territories
      1 CPC
      1 NDP
      1 LPC

      For the Prairies and Alberta, the numbers might be off because Nanos doesn't distinguish the two, so I simply used the same numbers for both regions, but that probably isn't right. It still gives you a general idea of the situation though.

      Delete
  13. It seems there was more audience than first reports and that Trudeau benefited from the very low expectations set by the attack ads.

    Backfire?

    Couldn't happen to more deserving folks.

    ReplyDelete
    Replies
    1. Maybe but, it could also be a temporary blip. I thought Mulcair bombed in the debate but, he only experienced minor loss. Mulcair looked over-rehearsed and couldn't improvise when he lost his place in his notes, he looked insincere, not genuine and very much like a career politician who's not in it for the little guy! After Trudeau's building an economy from the "heart out" comments the Conservative attack ads sure look justified; how does one build an economy from the heart out Justin? I'm sure no one knows including the leader of the Liberal party.

      Delete
  14. Now that looks more like a Forum poll.

    After that last one, I was worried they were losing their predictable character.

    ReplyDelete
  15. Eric,

    I enjoy watching the seat projections go up and down with the polls. I'm very curious to see which seats are the ones jumping around and between whom as the numbers change ever so slightly week to week in the projections.

    Is there anyway to puzzle this out on my own... or even better... way for you to flag it to my attention?

    ReplyDelete
  16. Dimitri Pantazopoulos!

    If ever you're talking to him, refer to him as "Mitch Pantazopoulos" one time; I promise he'll laugh out loud. He was once mislabelled as "Mitch Pantazopoulos" on a television newscast.

    Mitch was a different guy who worked in the same political office (*full disclosure - I also worked in that office*)

    ReplyDelete
    Replies
    1. Ha, maybe I'll try that next time.

      Delete
  17. Mainstreet says "Our Momentum Tracker continues to show the momentum with the Conservative party" ??? This despite their latest poll where the CPC dropped 7 points and the Liberals went up 4 points.

    ReplyDelete
    Replies
    1. I noticed that glaring inconsistency, too.

      Delete
    2. unless they meant negative momentum...?

      Delete
    3. Their Momentum Tracker is the same sort of thing as the Nanos Index, a blend of different factors. So the CPC dropped in voting intentions, but did better in their 'momentum' index.

      It is a good indication that their index might be showing something, but momentum is probably not it.

      Delete
    4. I believe it's looking at where undecided are leaning. The CPC and LPC are both doing better among "leaners" than their top line numbers suggest IIRC.

      Delete
    5. To me, that suggests only that the CPC might be able to regain the support they've lost. That's hardly momentum.

      Delete
  18. Eric, either I'm going blind or your riding projection is missing Newton. Or it's been moved to somewhere other than BC.

    ReplyDelete
    Replies
    1. It is called Surrey-Newton now.

      Delete
    2. Yah file it under "going blind." I didn't realize it was on the second image.

      Delete

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.