Monday, January 7, 2013

Angus-Reid gives two-point edge to Tories

On Friday, Angus-Reid released its latest federal numbers giving the Conservatives a narrow two-point advantage over the New Democrats in national voting intentions. They also included some numbers on the federal Liberal leadership race, showing that Justin Trudeau remains Canadians' favourite choice.
Angus-Reid was last in the field quite some time ago, between June 15-16 of last year. The Conservatives, at 35%, picked up one point since then while the New Democrats were down two points to 33%. The Liberals were unchanged at 19%.

The Greens were up one to 6% while the Bloc Québécois was unchanged at 6%. Another 2% said they would vote for other parties. None of these changes in support appear to be statistically significant.

Angus-Reid has been around now for a very long time and has had a lot of success, but their reports still leave something to be desired. Regional sample sizes should be a minimum, while unweighted samples would be great. It should not be a problem for Angus-Reid to do this, as their poll on Scottish independence that was released today includes a great deal of methodological information that the firm does not include in its Canadian reports.

The Conservatives led in Alberta with 55% support, followed by the NDP at 18% and the Liberals at 13%. Fourth in the province were the "others" at 8%, up seven points since June. That is an unusual result. The Tories also led in the Prairies with 57% to the NDP's 23% and the Liberals' 11%. That is a very low number for the New Democrats, the lowest since an Angus-Reid poll in May (going back 27 surveys). In fact, the lowest result that the NDP has registered since the 2011 election, 20%, was also by Angus-Reid. For some reason, the NDP can do very poorly in Angus-Reid polls in the Prairies.

The Conservatives were also ahead in Ontario with 36% to 35% for the NDP and 23% for the Liberals, and were up in British Columbia with 43% to 30% for the NDP and 16% for the Greens, a gain of nine points since June for that party. The Tories were also in front in Atlantic Canada with 38% and trailed by the NDP at 31% and the Liberals at 29%. The Greens dropped eight points to 2% in the region.

The New Democrats led in Quebec with 40% support, and were followed by the Bloc Québécois at 23%, the Liberals at 20%, and the Conservatives at 13%.
While the margin between the Conservatives and NDP was very small nationwide, the Conservative advantage in western Canada is crushing in the seat projection.

On the 338-seat map, the Conservatives would win 180 seats and a majority government, with the NDP winning 115, the Liberals 38, the Bloc three, and the Greens two.

The Conservative majority is really built in the West, thanks to dominating performances in the Prairies and British Columbia. This shows how essential strong NDP numbers in these two regions are if the opposition is to keep the Conservatives to a minority. It is a problem for the New Democrats if they win 83% of their seats in Ontario and Quebec, leaving 86% of the seats in the four western provinces to the Conservatives.

Angus-Reid included some approval rating numbers, showing Stephen Harper's approval rating virtually unchanged at 37%, compared to 54% disapproval. Thomas Mulcair managed 44% to 34% approval/disapproval, also steady, while Bob Rae dropped six points to 32% approval against 43% disapproval. Elizabeth May was unchanged at 35% approval. At 47%, she had the highest approval rating of the four leaders in British Columbia.

In terms of the Liberal leadership race, Angus-Reid shows that Justin Trudeau is still seen as a good choice by many more Canadians than his opponents, and he improves the Liberals' numbers dramatically. But the numbers also hint to a little stagnation, while Marc Garneau and, to a lesser extent, Martha Hall Findlay have improved their own numbers since June.
45% of Canadians think Trudeau would make a good choice as Liberal leader, up three points since June. Another 26% think he would be a bad choice, up two points. Those who are not sure stood at 29%, down five points. He is certainly gaining renown, but that renown is not improving perceptions (or hurting them).

But Marc Garneau was up eight points to 31% as to whether he would be a good choice, while his bad choice score dropped three points. He is still unknown to 56% of Canadians, but whereas Trudeau is not gaining any ground as he becomes better known as a candidate for leader, Garneau is making a good impression - even among those who had originally thought he would be a bad choice for leader. That is a positive sign for the former astronaut.

Hall Findlay's good choice score doubled to 12%, while 14% (-2) thought she would be a bad choice. Nevertheless, 74% of Canadians were still unsure. For Joyce Murray, Deborah Coyne, Karen McCrimmon, and George Takach, too many people were not sure of who they are or what they think of them to make their numbers very meaningful. It does seem, though, that about 1-in-10 respondents thought that their lack of renown made them a bad choice for leader.

But while Garneau does seem to be gaining some ground on Trudeau, he has actually lost ground on how his leadership would influence voting intentions. With Justin Trudeau as leader, the Liberals would take 42% of the vote in an election held today, compared to 26% for the Conservatives and 19% for the New Democrats. He takes nine points from the Tories and 14 from the NDP, almost certainly winning a large majority government on these numbers. The 42% is even an improvement on the 40% he registered in June's poll.

Garneau, on the other hand, would boost Liberal numbers to only 24%. That is four points less than he did in the June poll. The Conservatives take 32% and the NDP 27%. With Hall Findlay, she does not improve Liberal numbers at all. They stay at 19%, but the Tories drop to 33% and the NDP to 29%.

Why? The poll seems to show growing levels of frustration on the part of the respondents with the questions. Trudeau on the ballot sent one point's worth of decided voters to the "other" parties, boosting them to 3%. That rose to 6% with Garneau on the ballot and 8% with Hall Findlay. It is almost as if respondents were getting fed up with the questions. Why didn't these people say they were 'not sure'?

The hypothetical Trudeau numbers also seem to exasperate a lot of people in the commentariat. Did you know that François Legault was also polling really well before he launched the CAQ? If you didn't (or even if you did), I can find a few dozen people to tell you so.

That sort of approach misses the point. Will 42% of Canadians be willing to vote for Trudeau when he becomes leader? Probably not - unless he does extraordinarily well in the next few months. Undoubtedly, once a little reality meat is thrown on to the hypothetical leadership bones the numbers will change.

But what the 42% shows is that Canadians are open to voting for a Trudeau-led Liberal Party. It does not mean that their support is in the bag. But it does mean that Canadians will be paying attention, and that is something any political leader should desperately want. The question is what Trudeau would do that with that extra attention. Will he squander it and disappoint a large proportion of that 42%? That is what happened with François Legault and the other examples of new leaders who saw their stellar numbers plummet to Earth. It is a very high bar and that is why most new leaders don't live up to the hype. The challenge is to capitalize on as much of the opportunity that is presented as possible. Can Trudeau boost the Liberals to 42% in 2013? Probably not. But he can he boost them to 30%? That's where his candidacy becomes interesting.

30 comments:

  1. Those BC numbers look really suspect to me. The federal Conservatives are really not endearing themselves to BC lately.

    ReplyDelete
    Replies
    1. Agreed.

      Hence, Eric's aggregrate projections.

      Delete
  2. Good poll for Hall Findaly all the same, she's unknown by most yet the party would not lose any support with her leading them. So she has lots of room for growth, but could also become unpopular as she becomes better known I suppose.

    ReplyDelete
  3. Good article, Éric. Keep putting the pressure on the pollsters to release more data, and I hereby vow to stop bringing up François Legault, probably.

    Dom

    ReplyDelete
  4. Charles, in fact with the possible exception of Alberta, the Tory numbers across the country have an "aroma" !

    ReplyDelete
    Replies
    1. The CPC numbers in Ontario and Quebec seem reasonable to me. Both are in keeping with how the Conservatives have done in recent elections in those provinces. That said, I agree that the BC, Prairies and Atlantic numbers are suspect in this poll.

      Delete
    2. Quebec seems like a return to the new normal (post May 2011), but Ontario is a little surprising, with the NDP just 1% behind the Conservatives, and the Liberals returned to their depths in that region.

      Delete
  5. As to the BC numbers, the sample is likely to have been very small, realistically too small to give us a bell curve but more of speed bump curve .

    ReplyDelete
  6. Charles Harrison07 January, 2013 15:05

    Which seats did Angus-Reid give the Greens?

    ReplyDelete
    Replies
    1. Charles Harrison07 January, 2013 18:03

      Éric?

      Delete
    2. Seeing as they are both in B.C., I'll hazard a guess one of them is May's seat in Saanich--Gulf Islands.

      I wonder if byelections are taken into account in the projections. If so, maybe the other seat is an NDP-to-Green swing (never thought I'd write that, yikes) in Victoria.

      Delete
    3. Geez, give me a break. The vote in Victoria was an unofficial vote on the sewage treatment plant. Lizzie May will take the upsurge on a brown position. Just don't bring it up with her, her blood pressure can't take it.
      Victoria knows it was a by-election and a legit vote for the boring Galloway will not materialize in the next election.

      Delete
    4. I'm thinking Vancouver Centre? When the Green vote gets up to around 16-18% in BC they've taken that in the projection before.

      Delete
    5. Charles Harrison09 January, 2013 22:07

      I doubt it. Look at last election:

      Liberal 31.03
      New Democratic 26.04
      Conservative 26.04
      Green 15.44
      Libertarian 0.53
      Progressive Canadian 0.48
      Pirate 0.33
      Marxist–Leninist 0.11

      Delete
  7. Eric,

    As a proud Scot I appreciate the link to the independence referendum poll. I would be greatly interested if you ever did analysis of polls relating to this.

    Cheers,

    ReplyDelete
    Replies
    1. Not really my beat, but maybe in 2014 when the referendum will be held. That is setting up to be a calm year here.

      Delete
    2. Well if you get time it would be great. 2014 appears slow but, with minority governments in Ontario and Quebec you may find yourself unexpectedly busy.

      Thanks,

      Delete
  8. What were the 2 seats that you gave to the NDP in the Prairies?

    ReplyDelete
    Replies
    1. Most likely the two they currently hold in Manitoba: Churchill and Winnipeg Centre.

      With that sort of polling swing towards the Conservatives, there's no way the NDP would be picking up new seats in the projection. In fact, I'm surprised the projection didn't generate a CPC sweep of the Prairies/Alberta.

      Delete
  9. I wonder if any other candidates will attempt to make a late entry into the federal Liberal leadership race.

    I read that Chretien-era cabinet minister Martin Cauchon is considering a bid. He probably won't make too much of an impact as he was out of politics for a while, and as another Quebec Liberal. Plus, he was heavily defeated by Mulcair in Outremont riding in 2011.

    On the other hand, it would be good to have one or two more seasoned politicians entering the race. It would be good for the party and the country.

    ReplyDelete
    Replies
    1. I've never understood why a "seasoned politician" should be a good thing...

      Delete
    2. That's simple

      It's there to cover up the flavour of rot !!

      Delete
  10. "It is almost as if respondents were getting fed up with the questions. Why didn't these people say they were 'not sure'?"

    Having done some online polls this seems like a very likely explanation for the odd results. You need to complete the survey to get the points and the reward, but some of them are just too long and it is pretty easy to get 3/4s of the way through and decide to just click randomly to finish the survey.

    ReplyDelete
  11. Ive never really understood why some "Blue Liberals" are not just Tories?

    George O.

    ReplyDelete
    Replies
    1. Because the Paul Martins of the world do the job for 'em.

      Delete
    2. It's true Blue Liberals tend to be more fiscally conservative and pro-business, sometimes even more than the Conservatives.

      However, Blue Liberals are just like other Liberals when it comes to issues regarding social policy and national identity.

      Historically, the Liberal and Conservative parties were brokerage parties. Both parties were centrist and non-ideological. There were times when the Conservatives campaigned to the left of the Liberals, such as John Diefenbaker when he ran against Louis St. Laurent. It is also hard to see much ideological differences between a Red Tory like Joe Clark and a Blue Liberal like Paul Martin.

      Today, I could see Blue Liberals being most comfortable under the leadership of Marc Garneau. But even under Justin Trudeau, these right-wing Liberals will be served well. Trudeau has been running a centrist.

      - Maple

      Delete
  12. Éric, could you do some seat projections for plausible intermediate Trudeau-as-Liberal-leader scenarios?; ie if rather than the Liberals picking up 14 and 9 points from the NDP and Conservatives, what would the seat projections look like if they picked up something like 7 and 4.5?

    ReplyDelete
    Replies
    1. I've done some projections in the past for polls that had the Liberals at a high level of support - see the last few Nanos polls (you can find all of the polls I've written about by Nanos in the right-hand column).

      Delete
  13. One thing Nate Silver does a terrible job of is handling third parties (his predicted non-DR vote was off by 40% - I think he said 1.1 or 1.2% and they got 1.7 or 1.8%).

    On a related note:
    BC election - is there really a chance that the Green Party would get 0%? I'd say it is less than 1 in 1000. (Similar for the Conservatives, but I could see them falling apart more readily as they lack the long history of the Greens).

    For smaller parties, I suspect the standard error can be smaller than for larger parties (and this is a common error for polls). And the distribution is definitely NOT a normal distribution as you approach 0%. It is very hard for an established political party to get below 0.1%.

    ReplyDelete
    Replies
    1. Certainly, a result of 0% for either the Greens or the Conservatives is highly unlikely. It is a product of how far out we are from the election. The forecast does not assume equal probability for all possible results within the range.

      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.