Thursday, February 08, 2007

Party Pooper

Olaf, at The Prairie Wrangler posts the results from the first post-attack ads poll and the numbers are certainly interesting:
National horse race numbers give the Harper Tories their strongest score since a week before the 2006 election, 38% to 31% for the Liberals, 14% for the NDP and 8% for the Bloc.

• Tory strength in Ontario too: 40% to 35% for Liberals.

• Quebec numbers are a very different story: while the Tories (24%) are only a point below their 2006 election score and have essentially erased their bad summer scores in Quebec, the Liberals are at 32%, which is three points higher than in another house's poll a week ago and that party's highest score in Quebec since the Adscam audit. The Bloc, at 31%, continues to slide

As I already said at Olaf's, those Ontario numbers seem somewhat odd, given the other regional breakdowns of recent polls, which show the Liberals with a big lead. Did the attack ads have a positive result for the Tories? I want to see some more polling before I backoff my earlier "gloating" about those ads being a failure, but these numbers do give you pause.

To clarify, I agree that attack ads work, and I see the merit in the "seed" argument. I just don't think these ads were particularly strong, in fact they were surprisingly amateurish, given the warchest available. I also think these type of ads, prior to an election, are unprecedented, which makes predictability, based on past successes, all the more problematic. In addition, these ads are unique, in that the media is afforded complete focus, far different from the barrage of an election campaign. Were people's impressions of the ads formed through viewing, or were they aware, first, from the abundant coverage? I would argue the coverage was less than favorable, and allowed for plenty of partisan rebuttal, which furthers my opinion that the ads weren't effective. Having said that, I doubt many Liberal strategists are smiling at the Leger results.


Anonymous said...

Wait until a few more polls come out. but polls rise and fall, better to concentrate on doing good.

Having said that, this is the same polling outfit that registered a 3-pt conservative lead in Jan when all the other national polls were showing a tie or liberal lead. So I'd wait for others to get a better picture.

Anonymous said...

Knowing the structure of the survey and more specificity about the numbers would help a lot but at least on the face of how Wells reported it it's interesting.

Then again so is the Jan.25 Leger poll which shows quite different numbers that a few ads could barely begin to make a dent in.

The "which party would you vote for" question shows a statistical dead heat at Liberal 32% and Conservative 35%,

But the "do the Conservatives deserve to be re-elected" question shows a different dead heat at 45% no and 41 % yes with a whopping 11% deon't know/not sure.

It's here

There's no good news for anybody in an electorate as confused as Canadians are right now.

Except for anyone who's actively working to confuse them. It's good news for those people. Goddamn their eyes.

Steve V said...


"Confused" is an excellent way to characterize the current situation.

Monkey Loves to Fight said...

SES I think comes out next week so they seem to be usually the most accurate. If anything I think the polls show both the Liberals and Conservatives have 30% who they can count on voting for them for sure and then another remaining 10% who are up for grabs.

I do agree the Ontario numbers seem a bit fishy. I have seen ones showing the Tories only 5 points behind, but not 5 points ahead. That being said Ontario is not as monolithic as some think so depending on what part of the province the poll numbers come from and how they are weighted this could be a partial explanation.

I am less concerned about polls showing the Tories in the 40s in British Columbia or Atlantic Canada since the sample is under 100 so they are completely meaningless. You need at least 400 participants for the poll to even have some validity.

ottlib said...

This poll is just in keeping with how things have been going since the last election.

The Liberals are rock solid at about 32% +/- 2% and the Conservatives are jumping around like a yoyo.

I would say the increase in Conservative support, of 3 points, is a combination of survey error and the usual advantage that the incumbent party has in between elections.

That is the explanation for these estimates, not the ads. It is an unfortunate need of the media to find some kind of explanation for every poll result and in this case the Conservative ads are an obvious choice.

The Liberals can take solice in the fact the ads did not shake their support.

The Conservatives can take some solace in the fact they have a lead but they should still be concerned that their support is way too fickle for their own good.

Monkey Loves to Fight said...

I think both parties have a base of about 30% while a ceiling of 40%. At least one party makes a major blunder in the election I would expect both parties to get above 30% and neither to crack the 40%. In the case of Ontario, the Tories have consistently gotten between 32-38% in almost every election since confederation. Even during Trudeaumania they still got over 32% in Ontario while in 1988 they only get 39%. The only real change is the rural/urban divide seems to be becoming far more pronounced as they've picked up traditional rural Liberal ridings such as Glengarry-Prescott-Russell, Essex, Chatham-Kent-Essex, Renfrew-Nippissing-Pembroke, and Chatham-Kent-Essex while lost traditional Tory urban ridings such as Don Valley West, Etobicoke Centre, Oakville, Kitchener Centre, Kitchener-Waterloo, and London West.

This seems eerily reminiscent of all other English speaking countries since this divide also exists in the United States, Britain, New Zealand, and Australia.

I am not sure the exact reason for this divide, although my main hypothesis is rural voters are more traditional and don't like parties that promote major social change, while urban areas like progressive parties. In addition the demographics in rural areas are tilted in the Conservative favour since the median age is considerably older and as people age they tend to become more conservative. In fact in 1974, by a two to one margin, they were more voters going Liberal to PC than PC to Liberal yet the Liberals still increased their votes and PCs lost since the Liberals won significantly amongst first time voters while those who died between 1972-1974 were mostly Tories.