Friday, May 20, 2016

Provincial poll aggregations to be added


You may have noticed last week that I added a poll aggregation page for the upcoming provincial election in British Columbia. It is at the top of the right-hand column of the site.

Today, I have also added one for Nova Scotia and one for Ontario (they are a little further down the right-hand column).

My intention is to continue adding these pages until I have one for all 10 provinces. Links to them will be in the right-hand column, and they will be ordered from top to bottom according to the electoral calendar.

Though there are not always a lot of polls for every province, I've set these up so that they can be an easy reference for the latest polls. Each page features an aggregation (including a regional aggregation, if regional breakdowns exist), the monthly poll averages chart for the province, and links to all recent polls so that you can read them yourselves.

As time allows, I will add seat projections to these pages.

As I'm sure long-time readers have noticed, I have cut down drastically on original content here on ThreeHundredEight.com. The reason is simple — because I now work for the CBC, I do not have the time or energy to dedicate to original content here. I hope that these provincial poll reference pages (in addition to a federal one eventually) will give you reason to still come to the site frequently.

I will also continue to post snippets and links to my articles and podcasts for the CBC. That's the analysis you can expect to find here on a daily basis, whereas the provincial poll averages will be updated on a weekly basis as new polls are published.

I've also changed the font used at ThreeHundredEight.com to update the look of the site, which was getting very dated. A small tweak that I hope makes the site a little more readable. Obviously a complete overhaul of the design would be better, but that is not in the cards for the time being.

Comments on these changes are welcome!

The Pollcast: The state of the Conservative leadership race


With Michael Chong launching his leadership campaign this week, the race to replace Stephen Harper now has three contestants. What are their chances?

With a year to go before members of the Conservative Party cast ballots, Chong has joined fellow Ontario MP Kellie Leitch and Quebec MP Maxime Bernier in the marathon race. None of them, however, are seen as front runners — which is why they have launched their campaigns early in order to build up their profile and organization.

Can they use the time ahead of them to build a constituency large enough within the party to prevail? What impact might the upcoming party convention have on the race? And who will be the next Conservative to throw his or her hat into the ring?

Joining me to break down the race are Conservative insiders Tim Powers of Summa Strategies and Chad Rogers of Crestview Strategy.

You can listen to the latest episode of the Pollcast here.

Donald Trump and Hillary Clinton: not a coin toss yet


Last week, a new Reuters/Ipsos poll showed that Hillary Clinton and Donald Trump had drawn almost even in a general election match-up. The gap between the two candidates had dropped to just one point.

Headlines blared that the race for the White House was a toss-up. Clinton and Trump were neck-and-neck. Much ink was spilled.

The next day, Reuters/Ipsos was back in the field with their five-day rolling poll. This time, the gap between the two candidates had widened again to four points, a more conventional margin. The poll went mostly unnoticed.

Welcome to the fevered coverage of public opinion polling in the U.S. presidential election, which will culminate a mere 173 days from now. Expect polls which show a competitive race to get outsized attention compared to their duller counterparts.

You can read the rest of this analysis of the U.S. election here.

42 comments:

  1. These two have been pretty close for quite a while. The oddity is that the general public seems to dislike each at about the same amount. Weird !!

    ReplyDelete
    Replies
    1. The public dislikes Trump more, actually, that is born out in poll after poll of approval ratings and impressions.

      Delete
    2. Yeah I think so but the fact that the dislike is so high for both makes me wonder ??

      I'll vote Republican even though I don't like Trump or I'll vote Democrat even though I don't like Hillary.

      Or maybe I won't vote at all.

      Delete
    3. I think voting out of spite....such as a Sanders supported voting Trump, or a Cruz supporter voting Clinton seems unfathomable from a logical point of view....but obviously with millions of voters some will do just that....though I think a large majority of voters who are unhappy with either choice, will simply not vote at all....

      Delete
  2. Given the low predictive value of general election polls this far out, I'd say that any presidential election between any two candidates would be a coin toss at this point.

    The polls aren't yet useful for making predictions. The media is wrong to make a big deal out of the ones that show the race even, but you also shouldn't be pointing to the polls showing Hillary leading.

    They just don't mean anything yet.

    They especially don't mean anything because we don't know who their running mates will be. If Clinton were to pick Elizabeth Warren as her running mate, I would expect that to matter. If Trump selected Newt Gingrich, or Mad Dog Mattis, or Vince McMahon (nothing would suprise me at this point), I would expect that to matter.

    ReplyDelete
    Replies
    1. They have pretty good predictive value in the US. Their voting patterns are much more stable than ours.

      Delete
    2. They also tend to have much more regional or state by state polling than we do. But I do agree that numbers at this point are pretty meaningless other than what the start point will approximately be... not the finish line

      Delete
    3. Early polls are usually not predictive because the candidates aren't well known. Most presidential elections start off with at least one candidate who is not a household name until the autumn months preceding the election. Polls done before all candidates get significant media scrutiny are unreliable because people won't have made up their minds.

      Hillary Clinton and Donald Trump are two of the best-known people in America. It would be hard to find someone who didn't already have an opinion of both of them. Trump's poll numbers haven't changed much in the past 8 months much no matter what he has done or said, since most people already knew what to expect from him. I don't expect the polls to change much between now and November because most people have already decided (although the race might tighten up at the end if Clinton looks like the certain winner, as some people will then cast what thy see as a risk-free protest vote for Trump).

      Delete
    4. 8 months ago I would have agreed with Goaltender Interference, but that line of reasoning was thoroughly disrupted bt Donald Trump's pursuit of the Republican nomination.

      Everyone knew who Trump was at the start of his run, and his favourability numbers were terrible. Typically, a candidate improves those numbers by becoming better known (and finding new people who don't already dislike you). Using historical precedent, Trump's campaign was doomed.

      And yet he won. As such, I'm going to conclude that candidates that are well known at the start are different-in-kind rather than different-in-degree, so we have no historical basis by which to judge them.

      Also, Clinton has terrible net favourability numbers, too.

      Delete
    5. In a race with 17 candidates, Trump didn't need good favourability numbers to win the primaries. He could win early states with 25-30% support, which would never be enough to win a race with one or two other candidates. The GOP primaries had many states with winner-take-all states and other systems that favoured plurality winners, even where they were not close to a majority.

      It is similar to the presidential election in France in 2002: six minor left-leaning candidates ran for president and collectively got about 25% of the vote in the first round of voting (which is the closest thing that France has to "primaries"), while the strongest left-leaning candidate only got 16% of the vote. This allowed the anti-immigrant candidate (Le Pen) to beat all of the leftist candidates with less than 17% of the vote, thereby winning a spot on the final presidential ballot instead of the socialists. Le Pen was then crushed on the final ballot by the right-wing Jacques Chirac in the most lopsided vote in French history (82%-18%). Even though Chirac had horrible favourability ratings, all of the left-leaning vote went to the candidate they really didn't like instead of the anti-immigrant candidate that they despised.

      The US in 2016 is a bit different, in that right-wing voters in the US don't despise Trump as much as left-wing voters despised Le Pen. But I think you will also see a lopsided result, as the majority of GOP voters dislike Trump despite him having won the primaries, and so many will stay home in the general election.

      Delete
  3. They have such a dismal choice:

    One will be a horrible President and the other a Terrible President!

    Sound like Canadians voting between Harper and Mulroney??? LOL!

    ReplyDelete
  4. Don't know about the rest of you but I found Micheal Chong extremely impressive when he did his leader presentation. He's a "Progressive" to !!

    ReplyDelete
    Replies
    1. With Chong in charge, I'd likely vote CPC; with Bernier in change, I'd think about voting CPC; how Leith thinks anyone would vote for the CPC with her even part of the party, let alone leader, is quite beyond me.

      Delete
    2. Chong and Bernier are clearly the big tent conservatives in this race, and I don't see any others available. Leitch and Ambrose and Kenney and MacKay have bases of support that are too narrow.

      Delete
    3. I will agree re Chong. On Bernier I don't think so.

      Delete
  5. For the CPC leadership race, I think they would be well advised to look at the underlying trends.

    I think it is undeniable that Canada took a turn to the right during the early 80s and the next five prime ministers governed with a right wing slant (a liberal slaying the deficit, who would have thunk it?) with the exception of Paul Martin who opened up the spending taps and was "rewarded" with a minority win for his largesse.

    Yet, twenty years later Canada turned the page on that era when it purposely chose the candidate who promised mild deficits.

    The CPC, when should be mindful of this when selecting their new leader as it hints that a "we'll cut 100,000 jobs"-style candidate won't be successful and neither would be an antihijab-snitch line (see Hudak, Harper and Marois if you need confirmation).

    I'm not saying the CPC should select a liberal in CPC's clothing. Chretien wasn't a closet PC in the least, he simply tempered his decisions with the mood of the times.

    Someone like Michael Chong, with impeccable conservative credentials yet in tune with the mood of the times would be a powerful alternative to Trudeau. Someone from the old guard would practically assure a second term for JT.

    ReplyDelete
    Replies
    1. Hudak's problem wasn't the 100,000 jobs pledge. It was his total failure to explain how that would help.

      The winning side is the side with the better communicators. Almost every time. It has little to do with the minutia of policy. Kim Campbell was right when she said that election campaigns are no time to talk about policy, though that she said it marked her as a poor communicator.

      The Chrétien government didn't slay the deficit out of any ideological bent (though, Paul Martin did have conservative leanings on fiscal matters - just look at his banking regulations); it was largely a political manoeuvre in response to the strength of the Reform Party. Chrétien benefited tremendously from the split on the right, but if he governed in a left-leaning way he would have hastened the reunification of the Conservative Party.

      Delete
  6. "Hudak's problem wasn't the 100,000 jobs pledge. It was his total failure to explain how that would help."

    Twenty years before that Mike Harris didn't need to explain or apologize for it. What changed?

    "The Chrétien government didn't slay the deficit out of any ideological bent"

    Oh, I agree. It simply adapted to the times. Similarly, if the old style Reform guys stay in Mike Harris mode, my prediction is that they'll see a shellacking come next election.

    ReplyDelete
    Replies
    1. Mike Harris did a much better job of explaining what he was going to do, and why. And he was also up against a far less popular government.

      Harris's legacy was destroyed by Ernie Eves being an idiot after Mike left. People look back at the Harris era and see all the pain of the cuts but without any of the resultant benefits because Eves ruined it.

      I still proudly wear my Common Sense Revolution t-shirt, even though I've never lived in Ontario.

      There's also little need to be in old-style Reform mode right now. The financial health of the federal government is much better than it was in the early 1990s (largely because of the Chrétien-Martin budgets).

      Delete
  7. "There's also little need to be in old-style Reform mode right now. "

    Exactly, yet it seems Hudak never got the memo. Now, granted, Ontario is further in the red than the federal government; still my point is, the new leader of the CPC needs to put forward a platform for Canada in 2016 and not for Canada in 1996?

    Sometimes die hards within the party have a hard time updating their platforms. How many elections did Labour lost precisely because of this until finally Blair moved past their post war ideology mode they were stuck in?

    ReplyDelete
    Replies
    1. Right. Stephen Harper was probably the right guy for the job in 2004. He needed to look more reliably conservative than Paul Martin was, and Martin had some excellent conservative credentials after those 1990s budgets.

      Now the CPC needs someone who is way more hands-off on social issues, and more creative in terms of how to solve economic problems (though Trudeau might do a lot of that for them if he implements that guaranteed minimum income, which I think is a terrific idea).

      Delete
    2. Ontario, on the other hand, could use some shock therapy.

      And yet their finances are still arguably better than Québec's. Québec just keeps digging their hole quietly rather than making a big deal out of it.

      Delete
    3. Ontario owes about $100 B. more than Quebec 277 v. 181. So taking into account population both are roughly equally in debt per capita. However, based on the size of their economies Ontario's debt is approximately 38% of provincial GDP whereas Quebec's takes up 49%!

      Delete
    4. Exactly.

      Québec is basically doomed unless the feds bail them out at some point.

      Delete
  8. Mike Harris also proposed income tax cuts (his main promise in 1995), while Tim Hudak did not. Harris argued that balancing the budget had to share priority with restoring the Ontario economy (e.g. by facilitating personal investment), while Hudak placed priority on balancing the budget, even though budget cuts were a harder sell for Hudak, because he was not inheriting as big a deficit as Harris did. In this sense, Harris was closer in his general approach to that of Justin Trudeau, than Hudak was.

    ReplyDelete
    Replies
    1. Hudak's campaign was a screw up all around. Both he and Harper started with basic fundamentals that were on their side. Harper's side failed to propose a vision of Canada, sticking instead to the "just not ready" slogan.

      Hudak fell for the siren chants of sado-economists, proposing with glee 100K job cuts together with the infamous made up math of 1 million jobs. Interestingly enough bogus claims like this survive US elections unchallenged, while in Canada they were thoroughly debunked by all including well known right wing economists.

      Delete
  9. What an incredibly bad analysis.

    RCP tracks and averages all the significant polls comparing the national vote intention between Trump and Clinton.

    It was tied within the margin error for at least the last week. Now Trump is ahead in 3 out of 5 polls and ahead in the average.

    3 weeks ago Clinton was ahead by double digits. now it is exactly what you say it is not: A coin toss.

    Momentum would have Trump ahead by a large margin.

    Have you read and considered the explanation/rationalization from Nate Silver on how he let the anti-Trump punditry influence him into treating the polling data in a way such as totally making the wrong analysis?

    ReplyDelete
    Replies
    1. HuffPost Pollster, which seems to have a more complete record of polls, has Clinton ahead of Trump in 7 of the last 10, and by an average of a little under 2 points.

      Perhaps it is a coin toss now, with a lean towards Clinton. But this piece was written a week ago before the latest set of polls.

      Delete
    2. Polling average is not relevant and frankly it is the wrong metric to measure. U.S. Presidential elections are won or lost in the electoral college. The national vote is a nice aside but, any serious analysis should begin and end with an estimation of support in the electoral college. Clinton or Trump may be up by a dozen points or more in the national race but, it is where that support is located that is the important metric to measure.

      Delete
    3. You are going to actually put forth an argument that HuffPost should be considered to do a more accurate analysis than RCP???



      I have never ever heard any pundit, politician, or reporter use HuffPost Pollster as their source. On the other hand politicians and partisans on both sides frequently point to the RCP poll averages.

      Delete
    4. HuffPost Pollster has a more complete list of polls, and given the choice I prefer to look at all polls rather than just the ones RCP considers acceptable.

      Delete
    5. Derek - indeed. Stay tuned!

      Delete
    6. Cheers. Look forward to it!

      Delete
  10. Latest Nanos numbers out last night and despite/because of Elbowgate Justin up 3+% in approval !!

    ReplyDelete
    Replies
    1. Nanos polls over four weeks. This latest poll was 26 days before the elbow, 2 days after the elbow. It means nothing.

      The Ipsos and Abacus polls, however, mean something.

      Delete
  11. And what are they showing ?? No change ??

    ReplyDelete
  12. And this is what the site says
    Preferred PM
    Trudeau 53.8% +0.8
    Ambrose 15.6% -0.2
    Mulcair 9.0% -1.1
    May 3.6% +0.2
    Fortin 1.2% -0.1
    (Change b/w May 13 - 20/16)

    ReplyDelete
    Replies
    1. Meaningless poll. Trudeau is the only one likely to be still there come 2019. Mulcair or Ambrose becoming Prime Minister isn't in the card.

      Once the opposition parties have chosen new leaders, then we can actually see some useful data.

      Delete
    2. Well maybe Ira but if you think of the poll as being about the parties as well then it has more significance.

      Delete
    3. We have polls about the parties. This isn't one of them.

      Delete
    4. If you poll about the leader you are in essence polling about the party. Because they work together.

      Delete
    5. If you ask me which of those leaders would make the best Prime Minister, I'd choose Tom Mulcair, with Trudeau not too far behind. I like their temperment.

      But the NDP's polices are abhorrent to me. I don't want the NDP in charge. And I dislike what the Liberals are doing to our tax system. But there's no way I could point to Rona Ambrose and say I want her to be PM.

      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.