Monday, August 24, 2009

More on the Ipsos-Reid Poll

The details have been released in yesterday's Ipsos-Reid poll. Some interesting results.

I've already expressed my concerns with the accuracy of this poll, and the regionals haven't changed my mind.

To fill in the gaps, the NDP had 13% and the Greens had 11% in Ontario. The NDP had 8% and the Greens 9% in Quebec. At 26%, the NDP is competitive in Atlantic Canada and at 9% they are in trouble in Alberta.

The regionals in British Columbia and the Prairies are notable. In British Columbia, the NDP has dropped to 21% but more interestingly, the Greens are up to 14%. Has Ms. May's announcement of running in that province had an effect? In the Prairies, the Liberals polled only 9% - extremely low. And as you'll see from the other questions in this poll, Ipsos-Reid had trouble finding Liberal supporters in the Prairies. We know they're there, however, since recent polls have given the Liberals two or three times as much support as this one.

Conservatives lead in all demographics: 18-34 (27%), 34-54 (40%), 55+ (46%), males (41%), and females (37%).

This poll would translate into the following seat totals:

Conservatives - 154
Liberals - 83
Bloc Quebecois - 47
New Democrats - 24

Yes, that's right. One short of a majority. The Tory strength comes from Ontario, where they win 64 seats. Their results elsewhere are not out of the ordinary, though they do keep 9 of their 10 seats in Quebec.

Ipsos-Reid included plenty of other questions in this poll.

When asked whether they felt they had a clear idea of the policies the Liberals will enact in government, 44% agreed and 52% disagreed. I actually find that result surprising. People are aware of the policies of any party? Quebecers know about Michael Ignatieff's policies the most (54%), while the Prairies say they have no idea what he is talking about (72%). This is the first of the "we hate Mike" result in the Prairies.

As to what will influence voters, the perceived ability of a party and its leaders to follow through with its promises is tops, at 44%. Quebecers said this is the deciding factor the most, at 50%, while British Columbians don't care if politicians lie (36%).

The ability to handle the economy was second most important, at 37%. This is the top factor for British Columbians (46%) and Quebecers are least likely to consider this important (25%).

Finally, whether a particular party can form a majority is not a winning option - a further nail in the coffin in the Conservative strategy to ask for a majority. Only 14% of Canadians considered this the decided factor. Atlantic Canadians are most likely to be influenced by this, at 24%, while British Columbians are least likely (9%).

As to whether Liberals are ready to govern, 47% of Canadians agree while 49% disagree. Quebecers and Atlantic Canadians agreed the most (57%), while people in the Prairies disagreed the most (70%).

What's interesting is when you compare this result to the "should the Conservatives be re-elected result". Here, 45% of Canadians agree while 50% disagree. Looking at these two questions, the Liberals actually come out ahead.

Another question was whether the Liberals would do a better job handling the economic crisis than the Conservatives if they were the ones in power. 43% agree and 51% disagree, with the best/worst numbers coming in Quebec and the Prairies (guess which is which).

Now let's look at the Harper vs. Ignatieff numbers:

So Harper comes out ahead on the economy (48-40), world affairs (48-41), and finances (49-37). Ignatieff is in front on the environment (45-41). All of these results are surprisingly close. Harper's best numbers come in Alberta (65%-72%) and his worst come in Quebec (26%-32%). Ignatieff's best numbers are in Quebec (53%-60%) and his worst are in the Prairies (13%-21%).

This poll, because it is so different from the trends we've seen, will provide plenty of fodder for Canadian media and has already been the topic of choice in the blogosphere. I have a feeling, however, that the EKOS poll (and perhaps another poll) this week will contradict what Ipsos-Reid has found.

9 comments:

  1. Couldn't give me that one seat, could you :)

    ReplyDelete
  2. I'm telling you, no Tory majority without 40% in BC.

    ReplyDelete
  3. I actually see zero chance of a Con majority, unless Ignatieff gets arrested with a trunkload of coke during the election.

    ReplyDelete
  4. Hello Eric, I just finished reading your comment thread with Ad Rem, and don't blame you for being annoyed at his accusation of bias re: the Northern seats.

    Reading how you calculate your model, it makes good sense to include a shift in national numbers to the projections of the northern vote.
    I would like to offer a constructive opinion on a way in which I feel you could improve your modeling.

    Give the proportional national change the same weight as the 2008 election numbers.
    This way it can better reflect the incumbancy advantage, as well as the regions recent voting pattern.

    The proportional national numbers are valuable for sensing trends in the country as a whole, but the small population in the north may cause a change in the ROC to drown out any northern sentiment.

    Just a suggestion, as your model is your own, to calculate as you see fit.
    Instinctivly I find giving an equal weight would better reflect the northern trend, but have no-where near your expertise in determining these things.

    Keep up the good work!

    ReplyDelete
  5. -- "Give the proportional national change the same weight as the 2008 election numbers.
    This way it can better reflect the incumbancy advantage, as well as the regions recent voting pattern."

    Thanks for the advice, but if you look at the "Detailed Statistics" at the bottom of the page, you'll see that the 2008 and Proportional Change results already do have equal weight.

    ReplyDelete
  6. I see.
    I misunderstood your comment to Ad Rem when I thought you said you gave the national proportional change the most weight.

    One more quick suggestion if I may...

    In the leaders speak section I would find an english translation to french comments useful.

    In is a reflection of my anglophone bias of course, but my french is so weak that I can only get a gist of what is being quoted without pulling out the french/engish dictonary.

    A small thing to be sure, but I notice that all of your posts are exclusively in english, so I would imagine your readership would mostly understand it better.

    ReplyDelete
  7. --- "I misunderstood your comment to Ad Rem when I thought you said you gave the national proportional change the most weight."

    I believe you are right, and I was mistaken to say that it had the most weight. Your question made me look to see what weights I had given the elections and the proportional change.

    -- "In the leaders speak section I would find an english translation to french comments useful."

    I suppose you're right. Bilingualism is a de jure rather de facto reality in Canada.

    ReplyDelete
  8. Should be "rather than..."

    ReplyDelete

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.