Tuesday, January 17, 2012

EKOS highlights eligible vs. actual voters

Last week, EKOS released a series of poll results for iPolitics that were drawn from a survey conducted online in mid-December. Unfortunately, that makes these results slightly out of date but we can still take a look at them. I invite you to take a look as well at all of the other information contained in the expansive EKOS report, as it touches on much more than voting intentions.
This is the first publicly released report of federal vote intentions from EKOS since the May 2011 election, and the poll shows a much closer race between the Conservatives and the New Democrats than we've seen elsewhere.

The Conservatives finished with 31.4% support in this survey, compared to 29.5% for the New Democrats and 24.8% for the Liberals.

The Bloc Québécois stood at 6.7% while the Greens were at 6.1%.

But EKOS also filtered these numbers out according to who voted in the May 2011 election, weighing them accordingly. With those weightings, EKOS pegs Conservative support at 36.7%, with the NDP at 27.8% and the Liberals at 21.9%.

That is quite a big difference between the voting intentions of the general population and the voting population. It does not surprise me that there would be a disparity, though this is larger than I would have expected. This EKOS poll was taken at the same time as a Nanos poll (December 15-18), and EKOS's likely voter numbers are similar to what Nanos had found, though I believe Nanos reports the voting intentions of all eligible voters. Nevertheless, that Conservative support among actual voters would be far higher than it is among eligible voters jives with what we have seen in recent election results.

EKOS's main divergence from other surveys seems to be in British Columbia, where EKOS has the NDP leading with 35.9% to 29.1% for the Conservatives and 21.1% for the Liberals. Considering sample size, however, this is not problematic.

The poll confirms what we've seen in some other surveys taken at the end of 2011, notably that the race in Ontario is far closer than it was on election night, with the Conservatives at 34.6% to the Liberals' 31.9%. It also shows that the NDP has dropped in Quebec (34.4% to 27.4% for the Bloc) and that the Prairies are looking to be a real battlefield. In this poll, the margin between the Conservatives and New Democrats is tiny (42% to 38.7%).

Primarily because of the results in British Columbia and the Prairies, but also the close race in Ontario, these numbers would result in a small Conservative plurality of 118 seats. The New Democrats would take 103 and the Liberals 74, with 12 going to the Bloc Québécois and one to the Greens. If they could co-operate, the Liberals and NDP would be able to govern with a stronger majority than the Conservatives currently enjoy.

The Tories would take 12 seats in British Columbia, 26 in Alberta, 13 in the Prairies, 48 in Ontario, five in Quebec, 13 in Atlantic Canada, and one in the north with these numbers. In a 338-seat House of Commons, they would likely win 133 seats.

The New Democrats win 17 seats in British Columbia, one in Alberta, 10 in the Prairies, 23 in Ontario, 46 in Quebec, five in Atlantic Canada, and one in the north. In the expanded Commons, they would likely win 111 seats.

The Liberals win six seats in British Columbia, one in Alberta, five in the Prairies, 35 in Ontario, 12 in Quebec, 14 in Atlantic Canada, and one in the north. With the extra seats in the House, the Liberals would likely win 80 in total.

The EKOS results for the more likely voters would probably result in a much stronger Conservative minority, with the Liberals and Bloc Québécois making most of the gains at the expense of the Tories and NDP.

These are interesting results, and at the very least add another set of data showing that Conservative support has slipped since the election and that the NDP's support in Quebec is wobbly, though still enough to give them a majority of seats in the province. This may seem like an irrelevant point considering that the next election is more than three years away, but I'd counter that knowing how Canadians feel about the people and parties who govern them is worth knowing. And if the Conservatives increased their support to 60% or if they dropped to 10%, that would be a significant piece of information - even if it wouldn't change the make-up of our government for several years.

Note: The monthly federal polling averages chart has been updated to include this December 2011 EKOS poll.

13 comments:

  1. I'd be very interested to see if there is any change in the numbers following the Liberal convention, especially with thier newfound love for *all* things green. If the policy change results in a significant bump in the numbers, it may encourage them to hold the line on new pro-civil-liberty policies.

    ReplyDelete
  2. I don't think it's a newfound love, or a love at all even. Bills to decriminalize marijuana have been introduced by the Liberals in the past. It's just good policy.

    Eric, correct me if I'm wrong here, but you're using the all voters numbers for your projection here, right? I'd humbly suggest the likely voter numbers would be a better predictor of election numbers, as EKOS themselves have pointed out in the past.

    ReplyDelete
  3. "If they could co-operate, the Liberals and NDP would be able to govern with a stronger majority than the Conservatives currently enjoy."

    Eric it's also in line very closely with the Dec 2011 figures you have posted.

    Now what I've been noticing is that "comments" on various sites are distinctly showing a shift away from the CPC. A lot of real dissatisfaction seems to be appearing ??

    ReplyDelete
  4. Ryan, I did not see any regional breakdown for the likely voters, so I used the eligible voters.

    ReplyDelete
  5. Yah I just looked through that too. That's pretty bad on EKOS' part. Though I realize voting intentions aren't supposed to be the focus of this poll...

    I suppose you could scale the results but still... yeesh.

    ReplyDelete
  6. EKOS still has a long ways to go to clean up their act.

    We're going to need to see a full sample of likely voters (i'm guessing the MOE is higher here because the sample is smaller - higher accuracy, lower precesion).

    We're going to need regional breakdowns.

    Until then I see no point in pretending the CPC are at 31%.

    I see no point in running a projection on it or adding it to your model.

    It only serves the purpose of lowering your projection for the CPC.

    It was the same mistake you ended up making from '08 until the last election !

    ReplyDelete
  7. I think that CPC ability to turn out the vote is what gave them the majority. I would say the same thing with the Obama 2008 campaign south of the border( not to say he would have lost otherwise, but the amount of states he won and which states is astounding). Is there a new strategy brewing here to turn out the vote? i hope that the NDP learns from both campaigns in the next election.

    ReplyDelete
  8. I find it interesting that EKOS had likely voter numbers before the election which they didn't release. EKOS consistently showed lower CPC support than the other pollsters.

    I wonder if the other pollsters are already correcting for this, but doing it quietly.

    ReplyDelete
  9. IIRC the way Angus Reid constructs its online panels is based on previous election results, so it implicitly does correct for this? Maybe someone more knowledge knows?

    There's a bit of a built in likely-voter screen on different polling methods too (due to differences in response rates), which may be correcting it somewhat...

    IIRC Nanos is the only pollster still using live callers too, which may be why only them and Angus Reid got the results within their MOE last time.

    ReplyDelete
  10. Several firms use live callers, such as Ipsos-Reid, Harris-Decima, and some of the provincial or regional outfits.

    ReplyDelete
  11. The best surveys feature a very large sample of live calls where people are asked for various demographic information as well as questions on their voting history.

    From there calls are broken into sub groups and are weighted against the chances that they will vote in the next election.


    This is a very expensive process and involves some guess work.

    NO wonder nobody does it.

    ReplyDelete
  12. Oh right... I'm confusing myself here. Was Nanos the only one who doesn't prompt for any parties?

    ReplyDelete
  13. Goaltender Interference17 January, 2012 22:04

    Meh. The Bloc got a small boost from its leadership convention, the Liberals will get a small boost from their convention last weekend and the NDP will get a boost from their convention in a few weeks. Poll fluctuations right now are therefore pretty ephemeral. It will all wear off in the autumn when people get a better look at what the parties are like with their new leaders.

    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.