Wednesday, January 27, 2016

What do Trump, Clinton, Pallister, Couillard, and Trudeau have in common?

They are all leading in the polls!

- We'll start with the next election around the corner, the Democratic and Republican caucuses being held in Iowa on Monday. Hillary Clinton narrowly leads Bernie Sanders in the polls for the Democrats, while Donald Trump is narrowly ahead of Ted Cruz for the GOP. I wrote about the state of the race for the CBC here.

- In Manitoba,
 which votes in April, a new poll shows the Liberals are dropping in support, with the Progressive Conservatives taking advantage. The NDP still trails at a distance, tied with the Liberals. The poll is from Mainstreet Research and I also wrote about it for the CBC here.

- A poll by CROP conducted in Quebec flew under the radar, as the polling firm posted the results quietly to its website. Provincially, it shows the Liberals leading with the Parti Québécois dropping back. Both the CAQ and Québec Solidaire were up. Federally, the Liberals were way ahead of the other parties.

- And the latest federal numbers from EKOS show the Liberals enjoying a wide lead nationwide, with 47% support against 25% for the Conservatives and 16% for the NDP.

26 comments:

  1. The spin Mainstream put on their past record of accuracy in polling indicates a sense of complacency in their own techniques. They have a track record of being wide of the mark pre-campaign and early in a campaign before running back to the pack in their final polling. This latest poll seems to follow their past trend of overstating support for the Conservatives. Therefore I take their latest findings with a large pinch of salt.

    Mr Maggi talks about Rana Bokhari's visibility using as evidence the 41% of people who have yet to make up their mind about her. A leader's visibility and what voters think of them are two different things that are not necessarily related. Had Mainstreet asked a question about leader visibility Mr Maggi would have been correct to identify this issue as a challenge for Ms Bokhari, but he didn't so he shouldn't. He then adds
    "Interestingly while her support in Winnipeg is higher, the support she has in the rest of the province is actually stronger." I've no idea what this means.
    In terms of the party leaders approval/disapproval, Bokhari has a score of +13, compared with Pallister +21 and Selinger -31. If I was going to emphasise anything about how a party leader was perceived from my data, it would have been that.

    It is not clear from Mainstreet's reporting, how they are treating leaning voters. Their figures of 52 PC 20-20 Lib-NDP are said to be for Decided and Leaning voters. They then present seperate data for Leaning voters who are also described as Undecideds, PC 19 Lib 10 NDP 7. I would like to know if these voters are included in the main data as Leaning or Undecided.

    ReplyDelete
    Replies
    1. Anybody who has worked or visited Manitoba lately would not think 52% of decided voters off the mark at all. The animosity toward Sellinger is his palatable. He is in worse shape than Kim Campbell who had to carry Mulroney's record. Sellinger's record is all his own and people are looking forward to dishing out some punishment.

      Personally, I find your criticism way off the mark. You, Greg Pallister, have present no evidence to prove your allegation of bias toward Mainstreet. I would just say "put up or shut up" show some proof or data that demonstrates the bias through consecutive polls or their entire body of work or please hold your peace. As is, it appears you have an unfounded vendetta against Mainstreet.

      Delete
    2. You kind of expect commenters on here to have an awareness of polling generally, without providing details, so I suggest you read around your subject.

      Delete
    3. As I suspected you are unable to prove your unfounded allegations one kind of expect "commenters (sic) to have an awareness of polling generally" through the ability to provide verifiable data and one would hope the general knowledge of these "commenters" (sic) would be sufficient to provide data that proves their allegations of bias. Since, obviously you can not, I simply want to know why you feel obliged to "slander" a polling firm without sufficient (or any) evidence?

      I enjoy this site. Perhaps you do not. I am afraid your actions and "slander" against Mainstreet will have adverse effects for Eric. Analysing polls and critiquing rationale are par for the course but, one must do so within the realm of reason and science-through verifiable data not speculation. If you wish to critique their methodology you must demonstrate how Mainstreet's methodology diverges from ethical, industry or mathematical standards not make outrageous and unprovable claims such as: "The spin Mainstream (sic) put on their past record of accuracy in polling indicates a sense of complacency in their own techniques". Where is the example of spin? where is the analysis of their methodology to prove they are complacent? Where is the comparision with other polling firms? Or commentary from respected academics? None. You have no evidence you simply have a mental block accepting the Tories lead in the polls and are predicted to win the upcoming election. I submit the bias you "see" is your own and has nothing to do with Mainstreet6 Research's techniques.

      Delete
    4. Clearly you have not bothered to read about this subject as I had suggested and prefer to make wild personal accusations. As someone who enjoys this site and likes to comment on polls and pollsters, I will stick to this rather than engaging with others out to do otherwise.

      Delete
    5. You can't prove your accusations. You have not even quantified the supposed bias. People work for Mainstreet, why don't you write them and ask for an explanation? If none arrives it strengthens your argument, if one does it may answer many questions posed and inconsistencies you have noted. Surely, Mainstreet should have the chance to respond?

      Delete
  2. There are a few minor things that Mainstreet's data throws up which are hard to believe because either they go against accepted understanding or because they contradict other pollsters.

    Mainstreet polls the Greens at a whopping 9% among decided and leaning voters. Up 3% points over 2 weeks which is odd for a party that has zero profile. In 2011 the Greens only ran candidates in about a third of the seats and there is little indication that they are going to come close to matching that many candidates in 2016. If they manage it, then there will still be two thirds of the ridings where 9% of the vote will be up for grabs. If Greens are forced to vote for their second choice, Mainstreet says three times as many say they will vote Liberal than NDP. So where there is no Green running, that would give the Liberals a 2% point advantage.

    ReplyDelete
    Replies
    1. If the respondents say they're voting Green, Mainstreet should report that result.

      That those voters might not be able to vote Green shouldn't be Mainstreet's concern.

      Delete
    2. I'm not suggesting otherwise, merely commenting on the likely realities.

      Delete
  3. Good article for the CBC. Most polling analysis (Nate Silver aside) for the US race is painfully thin on insight. Usually just "wow'ing" over the horserace numbers but neglecting to mention important facts (The effect of race and religion in Iowa for example) or obsessing over national numbers which are of little import prior to Super Tuesday.

    My own basic analysis of the US race... if Trump and Clinton win their respective races in Iowa it's game over. Reasoning goes that Bernie's main constituency is white liberals and Iowa is one of the top three states where white liberals make up the majority of the DNC voter base (the other two being NH and Vermont). If Saunders loses there then the best he can do is walk out of NH with a "tie" before the electoral calendar moves on to states where he won't have such an advantage demographically speaking. On the other side of the aisle If Trump can pull it out over Cruz (darling of the evangelical tea-party set) in Iowa then he'll have an immense momentum boost heading into NH (where Cruz is a non-factor) and SC and beyond.

    If both those things happen then I'd like to congratulate the first female President Hillary Clinton. Trump would a historically unpopular candidate based on favorable.

    ReplyDelete
    Replies
    1. People keep saying trump would be unpopular, but he was unpopular among Republicans when he entered that race. And look at him now.

      Campaigns matter. Trump has shown an historically unusual ability to direct the media to his advantage. When he says something crazy, his opponents never have time to formulate responses before Trump has moved the topic on to the next crazy thing he said.

      The media (and the public) don't have much of an attention span, and Trump is using this to his tremendous advantage. He could well do the same in the general election.

      Delete
    2. I think that Donald Trump is a lot like Rob Ford. He's a populist that doesn't rely on facts or plausibility, but sentiment and visceral reactions. Trump might be able to win the Republican nomination, but he’s more likely to raise voter turn-out in the general election by polarizing voters.
      However, I find it very interesting that two of the top 3 candidates are second generation Cuban-Americans.

      Delete
    3. He's still "unpopular" among republicans. He has a positive net favorability rating but he's in sixth place by that measure among current candidates.

      Among non-republicans he has a -27 with independants, and a whooping -70 among democrats. He would be a historically unpopular candidate.

      Delete
    4. JF,

      I think you are jumping the Gun on a second president Clinton: Mike Hukabee and Rick Santorum are the last two Republican winners in Iowa-where are they now? Campaigning or at least doing charitable works with Donald Trump. I appreciate your logic and the demographic insight but, at the end of the day; if demographics were determinative Obama would have lost to Clinton in 2008. This election like every election is as much about the campaign as the candidates. Sanders has indicated he is in the race for the long haul by which time he may have pushed Hilary so far to the left she becomes unelectable. What if she loses New York?

      The system (and I am not defending it) was meant to foster these "small independent style candidacies", like Sanders and to give small states as much influence as large states: The system was made for surprises and to produce upsets against the establishment both Iowans and New Hampshiremen and women are known to have anti-establishment sentiment in their blood; Hence; Live Free or Die and the Iowa State Fair.

      Delete
  4. My biggest criticism of this poll is the survey samples. It has "Rest of Manitoba" with a bigger sample by nearly 100 people then Winnipeg, despite that in an election system where we have "one member, one vote", Winnipeg has a larger population then "Rest of Manitoba", we see this as Winnipeg has 31 seats while the rest of Manitoba has 27. So the PC lead is a bit over-stated. If anything, the PCs are probably closer to around 45-47% range. Mainstreet has been fairly out of ordinary, while all the other pollsters have been constant and consistent with each other

    ReplyDelete
    Replies
    1. Though it doesn't say so in the methodological statement - which I take to be an omission - the sample does seem to be weighted for regional populations.

      Delete
    2. Manitoba and all Canadian legislative assemblies and the House of Commons use First-Past-The-Post not One-Member-One-Vote.

      Delete
  5. And expect Bernie Sanders to win Iowa by a tiny amount. Public is starting to really understand Hillary is a "bought" item !!

    ReplyDelete
    Replies
    1. If Bernie Sanders were to win in South Carolina or another early diverse state, it'd be a big deal. Likewise, if he doesn't win Iowa, his outside chance looks a lot slimmer.

      Delete
    2. That seems unlikely. Bernie has an excellent chance in New Hampshire (though obviously that will change based on the Iowa result), but I'd say Hillary's about a 3:1 favourite to win Iowa.

      Not that it matters much. Iowa and New Hampshire are both older and whiter than the average state, and Bernie does better among older whiter voters. Bernie's not going to win any state with a significant black or hispanic population.

      Unless Hillary gets indicted for something during the campaign, she has the Democratic nomination in the bag.

      Delete
    3. Hillary will narrowly win Iowa, narrowly lose New Hampshire, overwhelmingly win South Carolina & Nevada.

      Delete
  6. We have heard for a long time now that the Liberals are picking up support almost exclusively from former NDP voters. Yet this poll is supposed to show that the PCs are gaining in a huge way from a drop in Liberal support? Doesn't make a lot of sense. It would be nice to get some non-Mainstreet numbers soon.

    ReplyDelete
    Replies
    1. I too found that odd. I also found it odd that there should be so much change between Mainstreet's fieldwork on 7th January and their fieldwork on 25th January. Outside of an election campaign you don't expect to find much change over just 18 days. Take the Greens, they have gone up 3 points in that time, despite having no profile to speak of. It is also odd that Mainstreet should poll so soon after their last effort, given how infrequently they have polled Manitoba.

      Delete
  7. Peter, as a fellow Liberal, I encourage you to consult Nate Silver's FiveHundredThirtyEight website for realistic information on what is happening on the ground in the US Primary races.
    Hope to see many of you bloggers at the Liberal Party of Canada B.C. Convention here in lovely Victoria March 11 - 13th, especially you Eric! No snow!

    ReplyDelete
  8. I haven't commented here since the federal election. I'm glad to see that you're still doing your always insightful analyses, Eric.

    Interesting numbers and I hope that we'll start seeing other pollsters around besides Mainstreet.

    My biggest problem with Mainstreet is that they always seem to overestimate Conservative support and underestimate Liberal support. I remember the huge outliner poll they had in early October that gave the Conservatives a nearly ten point lead. It wasn't until the final days of the campaign when their numbers finally aligned with every other pollster. Let's also not forget that their polls are commissioned by Postmedia, which is well known for its right-wing bias.

    I suspect that the numbers in SK, MB, and the Whitby-Oshawa by-election are closer than what Mainstreet is presenting. Hopefully we'll see other pollsters in the picture to determine if that's true.

    ReplyDelete
    Replies
    1. I agree.

      The comparative quotes from Mr Maggi are revealing. In early January he emphasized that he was surprised by the Liberals polling 27% in line with the findings of others. 18 days later he doesn't appear to be surprised at them dropping 7 points while the Tories put on 8 points. I wonder why he was commissioned so quickly to do another poll, and why it disagrees with his previous fieldwork.

      His latest figures giving the Liberals and the NDP 20% each are the best possible news for the Conservatives as it helps build a story of a scrap for second place and a divided centre-left vote, rather than one of the Liberals emerging as a clear challenger to the PC.

      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.