Brinksmanship, or, The Winnipeg Very Limited Strike


The Conservatives are on the brink of a majority! But wasn’t the Green Shift supposed to help? No, the Green Shift is really not helping. Here are some thoughts.

First, somebody is always on the brink of a majority when you’re the Globe and there’s an election around the corner. The paper used the same brink-of-a-majority language for its poll coverage in 2004, only then it was the Martin Liberals who were on the brink of a majority. And then they turned out to be on the brink of something else. I intend no prediction here, nor any particular criticism of the Globe; I’m only pointing out the natural tendency to want to ratchet up the drama when you’re about to pummel your readers with a month of campaign coverage. Have I mentioned this is the third Most Important Campaign Ever in four years? We are truly blessed as a nation. Or two.

Second, why do you suppose two news organizations timed their big, expensive national polls to come out on the morning of a Liberal caucus meeting? Because Canada’s top editors know Liberal MPs well, my friends! And they know there is no better way to get them in a panicky, mutinous mood than to suggest their next victory is not about to be handed to them on a silver platter.

What the Globe poll actually says is that the Tories are up a point, the Liberals down a point, from their standings in the 2006 election: not sunshiny news, if you’re a Grit, but survivable. What the CanWest poll shows is that “support” for the Permanent Tax On Everything (a confoundingly nebulous concept, “support,” but passons) is over 40% in every part of the country except Alberta and the Prairies, where the Liberals were not planning on major gains in any event. So if a Grit wanted to be cheerful, he could say the party’s main policy plank continues to be more popular than the party in much of the country, and that history has seen many worse hills to die on. (Well, a little cheerful.)

But isn’t it so much more fun, so much more cathartic, and so much more — what’s the word I’m looking for? Oh yeah — characteristic of Liberals to grab their favourite scribe by the elbow and whisper that The Leader’s Got To Go? Even though the campaign starts in five days and division is the surest route to collapse? I know many in this fine caucus are up to the challenge.


Brinksmanship, or, The Winnipeg Very Limited Strike

  1. So you think it’s easy, being the national newspaper?

  2. Some of these”polls” are all over the place but the dogs find a use for them.

  3. And in a regionally divided country with a Single Memer Plurality electoral system, where each party has certain strong regions, what good is a national poll?

    Will the liberals be looking at how they are doing in, say Interior BC? No. They want to know how they are in Vancouver/Victoria. If they are down 15 points in the BC hinterland, but up 5 in the urban areas that is GOOD news, despite the fact that they will show to be down overall in BC.

  4. “Brink” is a curious word, when you stare at it for a bit.

    In the online version, I didn’t see any breakdown by province; maybe it was different in the print edition?

    Anyway, here’s my math:

    With 124/308 seats, the Tories are currently 30 seats shy of a majority.

    They have 3 seats in NL, which presumably are toast, so they need to pick up 33 seats.

    If they cleaned the Liberals out of NB and NS (which I can hardly imagine), they’d pick up 12. Let’s say they win 8 of those, plus 2 in PEI.

    Reaching QC, they still need 22.

    Since they can’t win in Montreal, they can only hope to expand their base around Quebec City, with maybe the odd South Shore riding thrown in. What’s the best-case scenario there, +5?

    Skipping Ontario for a sec, MB right now is 8 Con, 3 Lib, 3 NDP. Let’s say the Tories manage to hang on to everything they have there and pick up 2 Lib and 1 NDP.

    In SK, there are two Liberal seats. Let’s say they win one.

    Nothing to gain in Alberta.

    In BC, it’s 18 Con, 7 Lib, 10 NDP, 1 Green. Say they take the Green seat, four NDP, and 3 Liberal seats. (I really have no idea of the situation on the ground, but this is a McCain-style analysis.) But they lose Emerson’s seat because he has a previously scheduled lynching. So the net gain I’m picturing here is +7.

    By this random reckoning, they still need 8 seats for a majority.

    So, I guess they could conceivably do it, if the Maritimes rejects the Liberals very strongly, they manage gains in QC, pick up urban seats in SK and MB, get a considerable number of additional seats in BC, and take a big chunk out of the Liberals / NDP in Ontario.

    Is this even remotely plausible?

  5. Jack, as long as “remotely plausible” is the test, I think the Conservatives could make much more substantial gains in Quebec. It depends whether people who’ve voted Bloc 5 elections in a row realize that’s no longer a novel thing to do.a small but real possibility.

  6. How does the polling and seat projection change when you exclude Alberta? I mean, isn’t it a bit of an advantage for the Liberals to concede Alberta, give the Conservatives a 28-0 head start, but, with the overwhelming Cosnervative polling support taken out of the equation, does that not give the Liberals a bit of a polling advantage over the Conservatives for the remaining 278 ridings?
    Does anyone follow me on this?

  7. In defense of those in the Liberal caucus with mutiny on their minds:

    1. The Conservatives were vulnerable at the time of the Bernier affair, but Dion had the Lib caucus sit on their hands because Stephane (I have the power) Dion probably thought his Green Shift would be the coup de grace.

    2. The Green Shift is a political mistake. For the true believers, look at the B.C. provincial numbers, where the NDP are leading Campbell and his Liberals for the first time in eons. Campbell raised gas prices on July 1st with his carbon tax. Sure it was only 3 cents a litre but Canadians, rightly or wrongly, regard cheaper gas as a birthright. Only 14% of Canadians list the environment as their main concern. Dion and May will be duking it out to carve up that little slice of the pie.

    3. Stephane (I have the power) Dion turns out not to have had the power. After dilly-dallying for months about bringing the government down about bringing the government down and bragging that only he, Stephane, had the power, it turns out Harper had some power too, and has decided, unlike Dion, to use it. Harper decisive, Dion anything but. Stephane, once again, looks impotent.

    There is more, much more. I, for one, forgive MPs in the Lib caucus who would be wondering out loud what is going on.

  8. You mean polls can be manipulated by the media? NO!

    Didn’t Harper suggest the same thing a couple of years ago, and wasn’t he chastised for it by a certain political pundit? Hmmmm.

  9. Baloneyman: I’ve often thought someone should start putting together TROC polls, like during the constitutional crisis years, but rather than remove Quebec from the subsequent results, take out Alberta instead. No offence meant to Albertans, I swear – it just seems easier to simply assume that the Tories will sweep the province, give them those seats as a baseline, and focus instead on areas that are more in flux when playing the seat projection game.

  10. Following up on Kady’s comment, it does look like there will be little movement in most areas of the country in the next election.

    There are two areas though, with some fluidity, and those are the area codes 418 around Quebec City and 905 around Toronto. The 418 is now mostly held by the Bloc, the 905 mostly held by the Liberals. That will be the battlefield. The Tories hold 10 odd seats in the 418 but there are an additional 20-22 seats there.

    If you’re the Conservatives, you gotta like fighting an election on territory held by your opponents. There are a few seats in B.C. that may now be in play, again, mostly Liberal seats.

  11. Jarrid: The only reason Stephan thought he had the power was because he made the msitake of “Taking Stephen Harper at his Word” with respect to the (ridiculous) fixed election date law, all within the context of Harper’s bellyaching and campaigning on the issue.
    But that, of course, is perhaps the greatest indictment of Dion as an astude politician: contrast any one of a number of aspect of Harper ‘pre-Prime Minister’ and ‘post PM’, and any neophyte political spring chicken ought to know that Stephen Harper’s word is as reliable as two-dollar watch. You cannot trust the guy to do other than mouth platitudes.

  12. An ideal poll would be to just poll the plus/minus 2% ridings where the difference between first and second was less than 2%. On a rolling basis, that would probably be more useful.

    Nanos/CPAC will probably do their daily tracking again. Given that Nanos is to polling what Jesus is to religion, I’d wait for that one.

  13. “Given that Nanos is to polling what Jesus is to religion, I’d wait for that one.”

    Nanos is a Liberal pollster, his spinning of his own poll numbers would make even the likes of James Carville blush.

  14. I think Green Shift is going to be a disaster for the Liberals and they know it. It is why Dion seems to be the only one trying to sell it. That Bob, Iggy and other ‘team’ members won’t touch the issue with a ten-foot pole tells us lots. Also, I have read a story or two about Lib caucus wanting to rework Green Shift into something more palatable.

    People aren’t buying Dion’s ‘plan’ to stop climate change, and end child poverty, just by tinkering with the tax code. People aren’t fooled by Dion’s decision to talk about income tax reductions but ignore the fact that other taxes are going to rise quite a bit and that there is no plan at all to reduce carbon emissions, unless hope and a prayer are a plan.

  15. Erm. Actually, Nik Nanos is a Conservative. His polling, on the other hand, is just plain uncanny, as far as prescience.

  16. I heard Nik Nanos was a communist with ties to both the IRS and the Chinese Olympic Committee.

    I lean Liberal and toy with voting Green. About the issues, one should note that environment is up from 14 to 15% in the Strategic Counsel poll.

    The BC numbers would be interesting. I don’t see how people are choosing Harper but the ridings in and around Vancouver/Victoria will be interesting.

    Dion is not aiming for Tory ridings so much as trying to suck in NDP voters. It’s a key dynamic and the ridings with tight NDP+Liberal vs Tory margins should be followed.

    The Liberals have an excellent “team” to sell their platform. I don’t see how Harper can use his ministers to help sell him as a good leader. Imagine all the local newscasts that can have McCallum, Iggy, Rae, etc appear live and then in the evening Dion gets covered nationally.

    Don’t forget, the Harper stereotyped voter watches the local news at 6pm.

  17. Actually, Nanos has been one of the most accurate pollsters in Canada when it comes to polling elections, and his most recent poll this week had the Liberals at 35, the Cons. at 33. I’m amused that the first instinct of some COns. supporters like Jarrid on here is to claim he’s a Liberal pollster; I’ve seen that elsewhere as well, Conservative supporters accusing pollsters who roll out #’s they don’t like as having the same “liberal bias” the entire Canadian media has, according to those folks.

    Decima, by the way, had it at 34-33 Liberals, and I think Ipsos Reid had it 33-31 Cons; all 3 of those pollsters having it a virtual tie within the MOE.

    The Green Shift hasn’t been hurting the liberals in those 3 pollsters numbers, in otherwards. So, I’d say to the Conservative supporters I’d be not be getting the victory drums out yet,

  18. Still not convinced that Harper wants an election. “The longer I am Prime Minister, the longer I am Prime Minister”, I forget who said that though…

  19. Paul – Thanks for being at least one journalist taking time to provide a more nuanced (and accurate) reporting. Just wished you would have went a little bit further in two areas:

    1. “The survey of 1,000 Canadians was conducted Aug. 25 to 31 and is accurate to within 3.1 percentage points, 95 per cent of the time.”

    This isn’t just a matter of statistical geekery (even if it is that too)….it legitimately means that all numbers reported ARE +/- 3.1….so the cons 1 point improvement and libs one point reduction creating a gulf of 8 points could actually be liberal increase of two points and conservative dip of 2 points resulting in the libs reducing the gap to 2 points… likewise it could also mean that the cons are actually at 40% and the libs are at 26%….

    all that too say that these 1000 person, nation polls’ value as barometers are at best even murkier then you suggest.

    2. It would be great if someone could provide a more nuanced reporting of leadership number as well. I know you didn;t go there I just wish you did. Here is the Globe’s ‘reporting’:

    “In head-to-head comparisons, Mr. Harper is found by Canadians to lead Mr. Dion in a number of areas, including who is better to deal with the United States and the economy, and who offers the most positive vision of the country. Mr. Harper is also seen by 53 per cent of Canadians as the most decisive leader, compared with 17 per cent for Mr. Dion.”

    If you click on the the graphic of results you see 4 leadership questions reported two falling in favour of Mr Harper two in favour of Mr Dion…Mr Harper’s are much more decisive (indeed subjecting these numbers to the same +/- 3.1 has one of Mr Dion’s potentially shift to Mr Harper and the other possibly become a tie) the

    We tend to get only the decisiveness numbers reported as a test of leadership. I haven’t tracked the value of each measure as an indicator on election day, but I think that the indicators “better at caring for someone like me” and “better at respecting the views of others” are perhaps not insignificant…TBD I guess. As far as decisiveness that could even be used against Mr Harper in a campaign (decisive about own interests re elections; decisive wrong way re reducing a food safety regs; etc etc).

    The point here is not to be a libs cheerleader…indeed I am also not sure you high numbers help more… it seems as a country we are pretty politically apathetic and perhaps the lib fans more so then the cons….i also think each ‘side’ is pretty ‘uncomfortable’ with the other taking a real hold of power so a reasonable hypothesis might be that to the degree one side is reported as ‘being up’ through these murky polls it might actually just motivate the other team’s base… TBD as well i guess.

    again…just wanted to celebrate your provision of nuance and cheer for some more

  20. When you’re right once it’s luck. When you’re right every time I’m not sure it’s luck anymore.

    That noted, I’d like to see Nanos’ polling on Dion’s carbon tax plans. As pollsters always point out the horserace numbers are the least revealing in terms of where the public is going. They move last.

  21. ” I don’t see how Harper can use his ministers to help sell him as a good leader. ”

    Good point. One of his ministers did recently come out and say about the new ad that ‘we just want Canadians to see the guy we work with everyday’. Of course, most Canadians wouldn’t want to spend much time with Harper’s ministers either, so their endorsement isn’t that compelling.

  22. How many refusals before getting the golden 1000 from across Canada? How many hang ups part way? What is the universe of the poll samples? Are VOIP / wireless / unlisted numbers included? How many answering the poll are goingt to vote or have any understanding of the questions? Soon after Green shift was announced a poll asked if people even heard of Green shift and out of the 1000 people answering only 3-5% heard of it. Non of these polls have any reliability.

    All the polls had the Alberta election as a close (possible minority government) for the campaign.

    The free polls published in the media are exactly worth what you pay for them $0.00

    Now to see the real polls taken by Harper and Dion and we would see why Dion has been running away for the last year….. just a guess

  23. Rudy, being “right” twice is not all the time. Nanos never altered his margin of error to less than one percent. It’s luck. Ask Zogby, who was hailed as a guru until he predicted a John Kerry victory.

    In fact, the theme of this blog post is about how polling is manipulated. It’s not a Bible.

  24. Jarrid, I would add 450 (Montreal suburbs), and, to hear tell from some, even 514 (Montreal island itself) maybe a little more fluid than many would believe. Certainly in popular vote movement, maybe in a seat or two or three. We’ll see, maybe.

    Kevin, I am with you there. I am not so sure we’re all going to the polls in October, either. This would be quite the head-fake, taking the media and the opposition for a ride. What will the opposition do, after whining that this is no time for an election, if everybody just goes back to work? Do they dare bring down the government after all the whining?

    The crow is in the oven, almost ready for serving, but count me as one who is not (yet) completely convinced PMSH visits GGMJ to dissolve.

  25. I’m reading in several places this is the 3rd election in 4 years. There were no elections in 2005 and 2007, so isn’t this the 3rd election in 5 years?

  26. I’m always amazed at how, when people bring what sounds like a higher level of political sophistication to poll analysis, they actually start saying more stuff that makes less sense. I’m not targeting anyone in particular. It’s a very general criticism. It must surely apply to me too.

    One big point: It is INSANE to compare pollsters based on their “predictive ability” when the election is six weeks away. Always remember: a poll tells you interesting things about the past and NOTHING about the future. People are going to change their minds. They just are. They will do so in response to… um… things that haven’t happened yet. No serious pollster claims his surveys can predict events in the future. I’m amazed when people debate the ability of such-and-such a pollster to do something the pollster would never claim to be able to do.

    A poll tells you, with considerable accuracy, how distracted people thought about something yesterday. It is an extraordinarily shaky proxy for how engaged voters will behave in mid-October, no matter who did the polling. And the main reason it’s so approximate isn’t that people are hard to poll, it’s that they change. This isn’t even an insight; it should be really, really obvious to everyone.

  27. To that end, would you say that poll *momentum* could be used as a predictor?

  28. Sure. Unless the momentum stops. Or reverses. If it does either, that will be in the future, and telephoning 1,000 people gives me no insight, none zero zip nada, about the future.

  29. I have to disagree Paul : Insanity is repeating the same behaviour over and over again and expecting different results.

  30. Because Canada’s top editors know Liberal MPs well, my friends!

    Verrrrry interesting.

    My take on the media is that they hope for a close horserace, but like to see conflict even more.

    I’m obviously partisan, and hope the close horserace is gone for good. But for entertainment value, it’s hard to beat the sight of the media pestering Dion over reports of dissention and disarray within Liberal ranks.

    Maybe you guys can harass him about his dual citizenship again ? That was fun to watch the first time, but reruns of good shows are welcome.

  31. Ryan,
    With regard to past elections and when they occurred.

    38th General election: June 28, 2004
    39th General election: Jan. 23, 2006
    40th General election: Oct. 14, 2008 (?)

    So that would technically be 3 elections in 4 years and 4 months.

  32. For what it’s worth … Minister Gordon O’Connor was on Dave Rutherford’s talk show about a month ago, and said that internal polls had the Tories at 36 %

  33. Jack Mitchell wrote:

    “Anyway, here’s my math:

    If they cleaned the Liberals out of NB and NS (which I can hardly imagine), they’d pick up 12. Let’s say they win 8 of those, plus 2 in PEI.

    So, I guess they could conceivably do it, if the Maritimes rejects the Liberals very strongly…”

    Greetings Jack, from the (real) east..

    To correct your math, there’s no chance of the Tories making gains here in NS. If anything changes it will probably be a Tory loss in Keddy’s riding. Thibault might keep opening his mouth but people don’t seem to get as offended as easily about things like out this way so he should be alright. Highly unlikely they pick up any seats in PEI. There only changes is one or two seats in NB.

    As for Dion’s “Green Shift”.. if he wants to come out on top of this one he better turn up the volume very quickly about the environmental plans of the other parties increasing costs to the consumer and giving nothing in return. The “Green Shift” doesn’t sound that great until you compare it to the other proposals.

    Anyway we’ll see how it goes.. Dion can’t run any worse a campaign than Martin. Maybe that we’ll give the Liberals some courage.

  34. Oh and MacDonald we’ll be out campaigning hard against Dion (Harper’s orders). He’s lame and won’t have much of an impact (here in NS anyway, not sure how the rest of the country will spin it).

  35. Well, if the Conservatives do get a majority, they better do everything they want with it because I’d hazzard to guess it’ll be the only one they get for a very long time. Once Grumpy and the gang are able to do fully what they please you can pretty much count on most of the country, save Alberta of course, will turn back to the Liberals and their new leader.

  36. As soon as I saw the poll had Ontario 35% Lib., 41% Lib., I dismissed it. If the general election sees that vote Harper may very well get a majority and try the neoconservative corporate tax cut mantra: raising revenue for capital intensive banks/oil and cutting employee intensity health-care. I don’t get why Conservative economists think more money to science-unfriendly oil is good and less money to publicly funded schools is bad.
    If Ontario is 40% of Canada, flipping the 6% yields 34.6% Conservatives nationally, and 31.4% Liberals.
    The irony is Conservative policies promote unemployment in bad times. China saves 35% of earnings and can ride out a US depression. We had a surplus once…

  37. MacDonald campaigning for Harper…….counter that with Danny Williams campaigning against Harper. Danny is smart enough to know that giving Harper a big Newfie goose egg on election – while the CONS still wins the government overall – is only wounding the bear. I wouldn’t be surprised to see Danny take his ABC campaign national.

  38. Sorry typos. Poll had 41% conservatives and Conservatives think less money to public school is good.

  39. Let’s just get this election going. From what I can gather from all these polls is that we are in the same place as we where when the last election ended.

    And who knows what will happen in this election.

    Let’s just get it on.

  40. Hazzard, they fully intend to:

    “When we come back with a majority then ALL bets are off.”
    — Gerry Ritz, Conservative Minister of Agriculture, February 13, 2008

  41. I never pay attention to Strategic Counsel polls. When compared to the results of other pollsters, they consistently favour of the Cons. The polling company I most respect is Nanos Research, formerly SES Research. The rolling poll Nanos did for the last federal election predicted the outcome within less than one per cent.

  42. Canadian politics can be so dull – we need a Dewey wins moment or, hey, someone exciting like ah, Sarah Palin.

    With what you say about editors, etc., so much for journalistic integrity. So, you just can’t trust the media.

  43. T. Thwim:
    Absolutely!! Majority parliaments are the least accountable, which is right up Harper’s alley.
    ‘Accountablilty,’ of course, means two things to Harper:
    1 – a vocalization that, if repeated often enough during an election, can cause gullible voters to support you,


    2 – (conceptually) something to be avoided like the plague

  44. Paul,

    I don’t have a favorite pollster and so I’ll just get back to one of the fun points you make here:

    it reminds me of a PQ caucus meeting when La Presse put out a poll showing Boisclair’s separatists were falling like Nortel stocks at the turn of the millenium.

    Boisclair lashed out at the meanies in the media for publishing a poll deliberately to muck up his meeting.

    It was the beginning of the end for the former Dear Leader.

  45. Well, if you don’t like the idea of picking a favorite pollster, you might want to check out the following website. It’s a rough equivalent of pollster.com for Canadian election polls.



  46. Strange things happening so far …

    Dion is gonna get tough on crime too. Or well at least “smart” on crime (as Martha Hall Finley used to describe it).

    Jason Kenny is out explaining the Greenshift.

    The two parties are swtitching platforms !

  47. Woah. That’s an interesting one. It looks like he’s trying to correct for polling bias by using the last election as a “zero-bias” pollster, and comparing where each firm had the various parties in relation to that, then extrapolating a consistent bias from it.

    Not sure how sound a methodology that is, but I’ve bookmarked it. I’ll be curious to see how well it matches up to the election results this year.

  48. Has CPAC announced they’ll be doing the rolling poll thing with Nanos again? Can’t find anything about it on their site… I sure hope so.

  49. ‘m from Alberta. After 40 years we have learned the hard way. “Never, ever, under-estimate the sheer stupidity of the soutern Ontario voter.”

    The election is wide open.

  50. Thanks Derek.

    Calling your fellow Canadians stupid really makes Albertans look good. It also suggests, no matter how rich Alberta might be, some of the people who live there might not yet be ready to wield the national responsibility such wealth entails.

Sign in to comment.