From the latest Nanos five-day regional roundups:

Top Conservative gains

1. Up 9 points in Alberta

2. Up 5 points in Manitoba/Saskatchewan

3. Up 1 point in the BC Lower Mainland

Top Liberal gains

1. Up 11 points in Rest of BC

2. Up 8 points in Northern and Eastern Ontario

3. Up 6 points in Montreal

Top NDP gains

1. Up 6 points in Toronto and the GTA

2. Up 6 points in Southwestern Ontario

3. Up 2 points in Rest of BC and Rest of Quebec

Highest regional Green support: 13% in Southwestern Ontario. Green support has held steady or declined in every region except Southwestern Ontario.


Top Conservative losses

1. 10 points in Southwestern Ontario

2. 9 points in Rest of BC

3. 8 points in Northern and Eastern Ontario

Top Liberal losses

1. 1 point in Alberta

2. 1 point in Sask/Man

3. Status quo in Rest of Quebec

Top NDP losses

1. 7 points in Alberta

2. 6 points in BC Lower Mainland

3. 2 points in Man/Sask


Conservatives lead in BC Lower Mainland, Rest of BC, Alberta, Man/Sask.

Liberals lead in Atlantic Canada, Montreal (tied with Bloc), Toronto and GTA, Northern and Eastern Ontario, Southwestern Ontario. Note the symmetry: Liberals lead east of Ontario-Manitoba border (except in most of Quebec, where the Bloc dominates), Conservatives lead west of that border.

NDP leads nowhere and is second only in Atlantic Canada.

Margins of error, even with a cumulative five-day sample, are between 6.0 and 9.7 depending on the region.

Filed under:


  1. The Conservatives must be thrilled to have gained 9 points in Alberta…. who says a majority is out the window?

  2. Two polling questions:

    1. What encompasses “Eastern Ontario” in this instance? I’m assuming it’s much more than just the Ottawa region though because the Tories currently hold 10 of the 12 Ottawa-area seats. And the only one anybody can foresee them losing is Orleans (Galipeau to Godbout).

    2. Why do the Ekos polls get virtually no attention (genuine question — and yes, I know Nanos nailed it last time)? Is it just because they don’t have a deal as an exclusive poller for a media outlet like the other companies do? (Nanos-CPAC; Decima-CP; Stragic Counsel-Globe/CTV: etc.). Or do they have some deficiency I’m not aware of?

  3. So possibly 7% of Alberta ndpers shifted to the Cons? who said irrational irony isn’t possible.

  4. I’m surprised by the Conservative lead in BC Lower Mainland, but maybe I shouldn’t be. I don’t think they split out BC regions in 2006 (i.e. only “BC” was listed I believe).

  5. Patrick: Ekos doesn’t have a media partner for this election, and some people are suspicious of its use of automated dialers instead of humans in call centres to do the survey. The automated technique has achieved good results in the U.S., but as this is the first time Ekos has used it, it’s not clear what to make of their results in this election.

    I’ve long put a lot of faith in Frank Graves and Ekos. It’s funny: they have a reputation as a Liberal-friendly firm, but they have consistently found a wider gap between Conservatives and the rest of the pack than other pollsters in this election.

  6. I find it interesting that the area of strongest support for the GPC is not on the left coast (BC) but in the small-urban and rural SW Ontario, which is also a bit of a manufacturing and industrial region. Too bad it won’t translate into seats. Ms. May should have run in Guelph. She would have had a huge portion of votes from the university.

    I think the damning thing about this summary is that the Liberals are the only party that haven’t lost ground in areas that matter. Indeed, CPC strength in Alberta is useless to them, making their voteshare less efficient (anyone tried factoring out Alberta from the national numbers?).

  7. The Liberals had a point in Alberta?

  8. Thanks Paul. I didn’t know that (re the automated technique). Will be interesting to see how their numbers stack up against the rest come next Tuesday.

    Are there any theories as to why/how automated polling could go wrong? Or I should say, why it would produce less reliable numbers? I suppose more people might be inclined to hang up on an automated caller as opposed to a real live person — but then we’d need a theory for how that translates to polling numbers favouring one party over another. I’m sure this is setting up a number of jokes for the folks on-line right now ;)

  9. Are there any theories as to why/how automated polling could go wrong? Or I should say, why it would produce less reliable numbers?

    My personal take is that certain kinds of people are more “rah rah rah”, and will use any and all opportunity to show their support. They just happen to be on one side of the political spectrum.

    And on the other side, you have people who just hang up because of the inconvenience.

    This kind of bias in sampling is just more likely for automated callers versus live callers.


  10. Indeed, support for the Greens here in SWO is a bit confusing, considering there’s nothing close to a ‘forest’ or wild animals for a couple hundred kilometres ;)

    It’s my impression that the less one encounters the natural environment the more they value it. Which I suppose makes sense in a way… but also not at all in another.

  11. Another thing is that it’s easier to get a valid result from automated dialing in the States where there are likely only 2 choices.

    I personally can barely remember the first three options on those voice-mail trees, and need to listen to the list at least twice.

    Also, there’s a built-in bias from the likelihood that they don’t rotate the order the names/parties are presented in, or at least it would be harder to truly randomize the presentation of 5 different taped questions.

  12. Thanks, SayWhaaat.

  13. As to the numbers, the Liberals being up 11 points in the Rest of BC doesn’t do much for them seatwise, especially if they’re taking Conservative votes (as it appears), since it puts the NDP into contention in a few more interior seats besides just Kamloops.

    However although I don’t like it that Nanos lumps Northern and Eastern Ontario together, the Libs gain in that one probably does mean they’ll take back both Ottawa-Orleans as well as Glengarry-Prescott-Russell, to answer Patrick’s first question. I can’t see their increase in that segment coming from northern Ontario as much because of the Green Shift, but maybe that’s just my rose-coloured glasses talking there.

    For the NDP to be up 6 points in Toronto and the GTA and up 6 points in Southwestern Ontario, on the other hand, probably does mean a few extra seats for them, especially if those points are coming at the expense of the Conservatives, while the Liberal vote in SWOnt is supposedly bleeding Green. SW Ontario is quite a big region, though … I suspect the Greens are stronger up in Huron-Bruce-Simcoe areas, while the NDP is stronger further south in the Windsor to Welland auto belt.

    The Atlantic numbers in all the major tracking polls have been pretty herky-jerky, but basically show a 3-way tie. You’ll have to call that one on a seat-by-seat basis.

    I almost never root for a Liberal to win (like REALLY almost never), but I am holding a kind thought for Ed Mahfouz in Nepean-Carleton who if we’re lucky will wipe the smirk off Pierre Poilievre’s face once and for all. For pretty much the same reason I’m also rooting for whichever Conservative is running against Liberal smarmy-pants Mark Holland. Getting rid of those 2 alone would improve decorum in the Commons by about 1000%.

    If I dare to dream, I do like the last minute trend-line for the NDP in Nanos’ Quebec numbers from the daily tracking poll (trending back up, today at 17%). I think their ads taking the shot at the Bloc are really powerful, and the maze imagery in particular was brilliant.

    Anyways, this is getting to the fun but scary part of the campaign, so here we go … !

  14. How can the NDP be not leading everywhere? Olivia Chow is adamant that her husband is the only one to lead us. I’ve seen it on TV. I’ve heard it on the radio. I’ve read it in the Star. I think I can hear it from rooftops in the downtown core. Clearly the pollsters are wrong.

  15. How accurate can polls from honing be when most Canadians refuse to answer polling and telemarketer calls? It seems easy to manipulate from the questions asked to answers given. My own father-in-law told a pollster to mind their own damn business. A secret ballet is still sacred to some.

  16. “How can the NDP be not leading everywhere? Olivia Chow is adamant that her husband is the only one to lead us. I’ve seen it on TV. I’ve heard it on the radio. I’ve read it in the Star. I think I can hear it from rooftops in the downtown core. Clearly the pollsters are wrong.”

    Richard, is there a reason you feel the need to be so nasty?

  17. What’s a secret ballet? Swan (Silent and Deep) Lake?

  18. As for the automated polling problems. A lot of people are probably being dishonest (Gilles Duceppe was, at one point, scoring 1-2% in Alberta and other prairie provinces) and machines can’t tell if they’re being lied to.

    I can never tell with these polls because I saw one showing the cons still at 40, and one with the libs-cons in a deadlock, and one with the ndp-libs tied. GAH!

  19. The Rest of BC gains for the Libs & losses for the Cons is almost all in the South Vancouver Island.

Sign in to comment.