Monday, April 14, 2014

Conservatives drop in EKOS poll

We have not heard much out of the federal polling world since the Quebec election took centre stage, and federal politics are likely to fade to the background again if Ontario heads into an election campaign next month. In the meantime, however, the latest poll from EKOS Research for iPolitics suggests the Conservatives have yet to rebound.

EKOS was last in the field January 22-27. Compared to that poll, the Liberals were up 2.6 points to 35.8% while the Conservatives were down three points to 26.7%. The New Democrats were down 2.5 points, while the Greens were up 1.9 points to 8.1%.

Only the drop in support for the Conservatives appears to be statistically significant.

The Bloc Québécois was up 0.3 points to 4.8%, and support for other parties was up 0.6 points to 2.7%. The number of undecideds were 10.2% of the entire sample, up 0.7 points.

This is generally the status quo, then, for federal politics since last fall. The number for the Conservatives is quite low, though EKOS generally has support for the larger parties lower than other firms due to higher Green and Others numbers.

In British Columbia, the Liberals were ahead with 32.2%, followed by the Conservatives at 25.8% (up 8.9 points) and the NDP at 25.7%. The Greens were at 13.8% in the province, their best result in the country.

The Conservatives were in front in Alberta with 53.7%, with the Liberals at 25.9% and the NDP at 9.6%.

In Saskatchewan, the Conservatives led with 46.3%, while the Liberals and NDP followed with 22.9% and 22%, respectively.

The Conservatives picked up 15.9 points to lead in Manitoba with 40.9%, while the Liberals dropped 19.6 points to 28.1%. The NDP was third with 21.1% support.

The Liberals led in Ontario with 40.4%, while the Conservatives fell by nine points to just 26%. The NDP had 21.1% support.

In Quebec, the Liberals were ahead with 33.8%, followed by the NDP at 25.2% (down 6.5 points), the Bloc Québécois at 18.6%, and the Conservatives at 15.4%.

And in Atlantic Canada, the Liberals led with 48.8% to 23.5% for the NDP and 21.4% for the Conservatives.

All in all, nothing too unusual in these regional numbers. However, the large Liberal lead in Ontario is somewhat out of the ordinary.

With these levels of support, the Liberals would likely win about 155 seats, putting them 14 short of a majority. The Conservatives would win 111, while the NDP would take 63 seats, the Bloc would hold seven, and the Greens would win two.

The Liberals do exceptionally well in Ontario, taking 73 seats with just 31 going to the Conservatives. They also win a majority of seats in Atlantic Canada with 23. The party takes the plurality of seats in Quebec with 34 and British Columbia with 14.

The Conservatives still do best west of Ontario, with 13 seats in B.C., 20 in the Prairies, and 30 in Alberta. The New Democrats take 26 seats in Quebec, with another 17 coming from Ontario and 13 in British Columbia.

EKOS also probed the approval ratings of the three main leaders, finding Justin Trudeau and Thomas Mulcair to have the highest approval scores at 41.3% and 39.5%, respectively. Trudeau had a higher disapproval rating at 34.2% to 23.7% for Mulcair, but Mulcair had 36.7% either not having an opinion or not responding. That compared to just 24.5% for Trudeau and 19.5% for Stephen Harper.

The Prime Minister had a very low approval rating of just 26.5%, compared to 54% disapproval. He had the highest approval rating among his own party's supporters, however, at 77.5%, compared to 77.1% for Trudeau among Liberals and 67.6% for Mulcair among New Democrats.

The numbers have yet to come unstuck, as Trudeau celebrates his first full year as leader of the Liberal Party. Over that time, he has not relinquished the national lead in the polls. But can he hold it for another 18 months?


  1. Still a long way from 169

  2. Hi Eric,
    Been a fan of your site for a couple of years now. Quick question: what (rough) methodology can I use to determine the likely level of support for a party in a particular riding based on a poll such as this one (or the aggregate total you provide)? Many thanks.

    1. The simplest thing to do would be to take the difference between the poll result in a province/region and the 2011 election, and apply that to a riding in the province.

      For example, if the Liberals are up 10 points in Ontario from 2011, add 10 points to their support in a given Ontario riding. It is very rough, but you could do worse. This is called uniform swing.

      I use proportional swing, which is a bit more complicated since you need to ensure that everything adds up to 100.

    2. Is it possible to explain proportional swing in simple terms with an example to the lay person like myself? Thanks again.

  3. 31 seats in Ontario for the Conservatives seems a bit high for 26.5% of the vote. In 2004 the Conservatives only got 24 seats with 31.5% of the vote, while the Liberals and the NDP are both outperforming their 2004 results in Ontario.

    And mathematical miracles of CAQ proportions would have to happen for the Conservatives to get 11 seats on 15% of the vote -- they are more likely to retain only the seats they have now.

    With a 9 point lead overall and a 14-point lead in Ontario, the Liberals would have to have the most inefficient vote in the universe not to pull off a slim majority.

    1. Remember Ontario gains 14 seats in 2015.

      Traditionally Ontario has been a conservative province hence the moniker Tory Ontario. At one time the grits could have counted on a dozen or so Northern Ontario ridings today most of those seats lean NDP, which explains I would think how the NDP can pull off 17 seats with 21% of the vote.

    2. Traditionally, Ontario has been Liberal federally and Tory provincially.

      The Big Blue Machine dominated the provincial level, but those Tories were urban centrists, while the provincial Liberals relied on pockets of support in rural and Northern Ontario. The provincial Liberals were also a third party rump during the 1970s.

      On the other hand, Ontario leans more to the Liberals federally. Which allowed Pearson, Trudeau, Chretien and Martin to have their victories. The Liberal coalition has usually been Ontario+Quebec+Atlantic. In 1988 and 2006 the Liberals "won" Ontario while losing the election.

      That being said, the Tories will have an incumbent advantage in some of those close Ontario races that could yield them a better result than what projections say.

    3. Traditionally, Ontario has been Liberal federally and PC provincially.

  4. Using the 308 map (yup, still holding up on the changes to the new map, I'm lazy...), I get:

    135 LPC
    99 CPC
    67 NDP
    6 BQ
    1 GPC

    Region by region, it would give me:

    22 LPC
    6 CPC
    4 NDP

    36 LPC
    25 NDP
    8 CPC
    6 BQ

    57 LPC
    30 CPC
    19 NDP

    18 CPC
    5 LPC
    5 NDP

    25 CPC
    2 LPC
    1 NDP

    British Colombia:
    12 CPC
    12 NDP
    11 LPC
    1 GPC

    2 LPC
    1 NDP

    I have never seen such an inefficient vote as the Liberal's in BC. They are leading province-wide, but they average around the late 20's in most riding, making them unable to win many, which is the opposite of the NDP and CPC who have a concentrated vote with very high ridings and very low ridings, giving them more seats (according to my simulator anyway) than the LPC, even if they have a lower provincial score. Interesting indeed.

    1. Most of the Liberal lead in BC is held in Vancouver proper and to a lesser extent Victoria. The Liberals do this every election cycle in BC the rack up large leads in the polls only for it to halve on election day.

  5. Hi Thierry,
    Thanks for your post. I live in Alberta. The 1 NDP seat is obvious (Edm-Strathcona), but where do the 2 Liberal seats come from? Calgary Centre and Edmonton Centre? Thanks.

    1. I don't know if those ridings changed since they are from the 308 map, but they would be:


      And you are right about the NDP riding.

  6. The Harper Government seems intent on squandering their political capital The new "fair" elections Act is the latest example. If it passes as is I'm one Conservative who won't be supporting them in 2015.

    We can survive four years of Trudeau if Necessary to get rid of Stephan Harper.

    Nice to Rob Anders out in Calgary.

    1. I think most people feel it is reasonable to present ID before one votes. Vouching (whatever its merits) is a 19th century mechanism meant for a time when Canada's population was rural and government rarely if ever issued ID. Being able to present 1 piece of photo ID and another with ID or statement with your address is a very small burden. Most Conservative probably think being able to present identification reasonable, you can't drive without ID, you can't fly without ID, you can't buy alcohol or cigarettes without ID, or go to a licenced restaurant or bar without ID. If a person can't provide themselves with photo ID and a bill or statement with one's address for an election in 18 months time maybe they shouldn't vote?

  7. Something tells me that the Bloc won't have any seats left at all by the time of the next election. Their hardcore volunteers are devastated after the PQ defeat. The PQ and Bloc finances are exhausted thanks to the Quebec election and Harper's changes to party financing. They're also leaderless at the moment. While I suppose one of the PQ's defeated candidates might try for the leadership, I think it unlikely anyone significant will attempt to revive the Bloc. I can't see them going into another campaign in less than 2 years with any enthusiasm.

  8. EKOS polls always overestimate Green support relative to other polls, cutting into the support for the others. I like that EKOS asks some good questions about leader approval rating. It's interesting that Mulcair and Trudeau have the same approval rating, but that Trudeau has a higher disapproval rating. It seems a number of people still do not know who Mulcair is. That would indicate that the NDP has the most to gain from a good campaign.

  9. Hey Eric, I have a question. Do you have a quick link somewhere for the rankings of the pollsters at the federal level? I took a quick look at the methodology, but didn't see one. I could easily have missed it (there's alot of stuff!).

    If on the other hand it were a bit of a trade secret, I could understand that as well. I believe it is at the heart of why your model out-preforms individual polls. I was always interested in this info (is Angus Reed still on top?), and find it valuable context.

    Also, Have you secured the 338 web address yet?

    Regards, and congrats on your excellent QC election performance. The CAQ seat surge could not have been accurately predicted, as a late surge in polls can either attract or frighten people, and QC politics is notoriously volatile.

    Thanks for all the hard work. Great tool for political junkies.

    1. The ratings are a 'trade secret', primarily to avoid needlessly antagonizing the pollsters.

      The site will remain at and the name will stay the same as well.



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.

Note: Only a member of this blog may post a comment.