Canadian Political Polls

[Bryan]

It's also interesting to me that Nanos does not give the respondents a list of parties. In my mind, while I know he was most accurate in 2006, this is silly. In the booth, I get a list of parties. I'm not sure why you wouldn't give a list here.

That is odd. Not sure it means anything either though. My pencil always hovers over one of the fringe parties before I mark my real vote. Never having been given the option, I never would have named this fringe party as my choice, but even when it was there, I still didn't vote for them even if I did think about it for a second.

As much as Nanos is usually my "go to" , their samples are smaller, their margin of error larger, and their recent numbers a little hard to believe. I'm begining to wonder if in past elections that they might have been leading public opinion rather than reflecting it.

[Jobu]

That is odd. Not sure it means anything either though. My pencil always hovers over one of the fringe parties before I mark my real vote. Never having been given the option, I never would have named this fringe party as my choice, but even when it was there, I still didn't vote for them even if I did think about it for a second.

As much as Nanos is usually my "go to" , their samples are smaller, their margin of error larger, and their recent numbers a little hard to believe. I'm begining to wonder if in past elections that they might have been leading public opinion rather than reflecting it.

Yeah, I can't really put much stock in a poll that has margins of error of about +/- 10% in more than one region and bases its results in interviews of less than 100 people in certain cases. I mean, Dion's and Harper's leadership scores will go up and then down 10-20 points overnight sometimes, on no news.

Different pollsters are more or less accurate in different elections for a whole host of reasons. I think Zogby or one of the other pollsters nailed the 2004 election, but was way off in 2000. Nanos gets the benefit of the doubt from most for his recent success, but in this businses, particularly given his overnight swings, I'm not sure any of the polls we've seen are helpful.

[White Doors]

I think that's part of the thing. Many countries, including the US and much of the EU, are beginning to say that they won't trade with countries that don't reduce their pollution (including carbon dioxide). We have to do something or we'll be left behind. I still really don't like the Green Shift, but so many seem to think that its good policy.

Yeah, except Carbon Dioxide isn't a pollutant.

[White Doors]

That is odd. Not sure it means anything either though. My pencil always hovers over one of the fringe parties before I mark my real vote. Never having been given the option, I never would have named this fringe party as my choice, but even when it was there, I still didn't vote for them even if I did think about it for a second.

As much as Nanos is usually my "go to" , their samples are smaller, their margin of error larger, and their recent numbers a little hard to believe. I'm begining to wonder if in past elections that they might have been leading public opinion rather than reflecting it.

Their reasoning about not naming the parties is probably that they have to say them in some sort of order and are worried about the impacts of which one they mention first and last etc. If people don't know a name of a politcal party then they are unlikely to vote for them or vote at all.

just a thought

[normanchateau]

http://www.ekoselection.com/index.php/2008...october-8-2008/

Latest EKOS poll has lead up to 11 from 9 and NDP closing in on LPC, and perhaps more significantly the CPC having a heavy edge in voters who will actually vote:

CPC 35

LPC 24

NDP 20

Green 11

Bloc 10

[OTTAWA – October 8, 2008] – The Conservatives continue to maintain a significant lead nationally over all the opposition parties, according to the latest EKOS tracking poll, and have a significant lead with some of the most important groups of voters, but they are fighting a series of very different regional battles.

“Across the country, the Conservatives hold significant leads among many of the demographic groups most likely to vote, such as seniors, men and baby-boomers,” said EKOS President Frank Graves. “There is nothing in this picture that suggests that the Conservatives have been knocked off track from winning government again, by either the leaders’ debates or the international financial crisis.”

“We see no sign of a Tory collapse or a Liberal surge. It does seem that some voters have been drifting between parties in recent days, perhaps because none of them inspires much enthusiasm with the electorate,” he said. “There is some tentativeness and instability, and the gap with the Liberals was narrowing for a few days but the Tory lead, if anything is widening now, not shrinking.”

The Ekos poll the day before that had CPC at 34% and the day before that at 33%. In other words, Ekos has CPC trending up. If this trend continues at this rate for another five days, it suggests a possible Harper majority.

Why should we trust the Ekos numbers more than Nanos or Harris/Decima? Ekos has a far, far larger sample size than either of the other two polls. A few days ago I anticipated a reduced Harper minority government but a majority remains a distinct possibility.

[kengs333]

The Ekos poll the day before that had CPC at 34% and the day before that at 33%. In other words, Ekos has CPC trending up. If this trend continues at this rate for another five days, it suggests a possible Harper majority.

I wonder what "margin of error (+/-) 1.7" means?

There has been a noticeable downward trend for the Cons the last week or so, this only seems to suggest that it is stabilizing, not going back up. I think the 30-35% range is what the party could always have counted on, so it means that they really haven't made any gains in this election or haven't screwed things up enough to shift support to other parties. So its a matter of what portion of the remaining 65% the other parties will get, and as we all should know, there is a good chance that there will be an election day shift to the Liberals. The Greens always poll high, getting half on election day; many NDP supporters will vote Liberal in ridings where the Liberals have a chance of winning. If the Greens are at 10 or 11%, they should get at most 6% on election day. The NDP will probably get 15 to 18%. The Bloc will drop to about 8%. So my guess is:

Lib 35%

CPC 34%

NDP 16%

Green 6%

Bloc 8%

Other 1%

[White Doors]

I wonder what "margin of error (+/-) 1.7" means?

There has been a noticeable downward trend for the Cons the last week or so, this only seems to suggest that it is stabilizing, not going back up. I think the 30-35% range is what the party could always have counted on, so it means that they really haven't made any gains in this election or haven't screwed things up enough to shift support to other parties. So its a matter of what portion of the remaining 65% the other parties will get, and as we all should know, there is a good chance that there will be an election day shift to the Liberals. The Greens always poll high, getting half on election day; many NDP supporters will vote Liberal in ridings where the Liberals have a chance of winning. If the Greens are at 10 or 11%, they should get at most 6% on election day. The NDP will probably get 15 to 18%. The Bloc will drop to about 8%. So my guess is:

Lib 35%

CPC 34%

NDP 16%

Green 6%

Bloc 8%

Other 1%

So you are predicting a minority liberal government?

[Vancouver King]

It's hard to argue with success so here are the newest Nanos numbers:

CPC - 33%

Libs - 29

NDP - 20

Bloc - 10

GRN - 7

Taken with yesterday's results it appears the Tory slide has stabilized at levels which at best will yield them another minority with reduced seats.

Edited October 9, 2008 by Vancouver King

[PoliticalCitizen]

Canadian Press / Harris-Decima:

Conservatives - 32%

Liberals - 27%

NDP - 19%

Green Party - 12%

BQ - 8%

http://www.harrisdecima.com/en/downloads/p...ses/100908E.pdf

[kengs333]

So you are predicting a minority liberal government?

Slim minority either way, but probably for the Conservatives...

[Vancouver King]

Canadian Press / Harris-Decima:

Conservatives - 32%

Liberals - 27%

NDP - 19%

Green Party - 12%

BQ - 8%

http://www.harrisdecima.com/en/downloads/p...ses/100908E.pdf

Dion should shake this race up again. My suggestion: Next campaign stop he should ask whether Harper will follow suit if the next American president declares martial law in 2009.

[M.Dancer]

Dion should shake this race up again. My suggestion: Next campaign stop he should ask whether Harper will follow suit if the next American president declares martial law in 2009.

If he needs dim witted advice I'm sure he knows already where he can find it.

[Vancouver King]

If he needs dim witted advice I'm sure he knows already where he can find it.

Yes, those geniuses running the Tory's Quebec campaign will do just fine.

[normanchateau]

Ekos, which has by far the largest sample size, shows the following trend in Conservative voting intentions.

October 9 36%

October 8 35%

October 7 34%

October 6 33%

Ekos seat projections released October 10th are:

CPC 152

Liberal 60

NDP 39

BQ 57

http://www.ekoselection.com/index.php/2008...ctober-10-2008/

Ekos concludes:

"We have not taken into account the fact that the Conservatives are stronger among demographic groups that have a higher propensity to vote, that their voters are more likely to say that they are “absolutely certain” to vote and that this is an important election.

In sum, we think if anything this projection may somewhat understate Conservative strength."

Given that the above projections underestimate Conservative's propensity to vote, a Conservative majority and a BQ official opposition are a distinct possibility next week.

[Slim MacSquinty]

So the comments mean that when people say they are voting conservative, they really mean it and that conservatives are reliable and will actually vote.

I just couldn't help myself.

[jdobbin]

Given that the above projections underestimate Conservative's propensity to vote, a Conservative majority and a BQ official opposition are a distinct possibility next week.

If we can trust the automated phone call system. This will be the real test. It is the first Canadian election to attempt a rolling poll in his matter.

[Vancouver King]

CTV - Strategic Counsel, 'Battleground 2008'. This is interesting movement.

http://www.ctvbc.ctv.ca/servlet/an/local/C...BritishColumbia

This is completely counter to the Ekos numbers above and, on balance, some of the best polling news the Liberals have had this campaign.

[Jobu]

CTV - Strategic Counsel, 'Battleground 2008'. This is interesting movement.

http://www.ctvbc.ctv.ca/servlet/an/local/C...BritishColumbia

This is completely counter to the Ekos numbers above and, on balance, some of the best polling news the Liberals have had this campaign.

This polling is up to Oct 8th and has ridiculous margins of error.

I don't think anyone takes Strategic Counsel or Donolo seriously, as much as I have liked some of his numbers in the past.

[Vancouver King]

This polling is up to Oct 8th and has ridiculous margins of error.

I don't think anyone takes Strategic Counsel or Donolo seriously, as much as I have liked some of his numbers in the past.

Picking and choosing the numbers you like seems to be a problem with you.

[Jobu]

Picking and choosing the numbers you like seems to be a problem with you.

It's not about "like", it's about margins of error and track record. EKOS and Nanos have the track record, far more so than Harris Decima or Strategic Counsel. EKOS has way higher sample sizes this time around, too.

[Vancouver King]

It's not about "like", it's about margins of error and track record. EKOS and Nanos have the track record, far more so than Harris Decima or Strategic Counsel. EKOS has way higher sample sizes this time around, too.

You are half right. Nanos had the 2006 result predicted accurately to within one one-hundredth of one percent of the actual outcomes for all 5 major parties. No one else, including EKOS, was close.

Then again, anyone who would use the bandwidth yesterday that you did extolling the liquidity of our banks only to wake up to today's bailout can't be expected to demonstrate much objectivity.

[Jobu]

You are half right. Nanos had the 2006 result predicted accurately to within one one-hundredth of one percent of the actual outcomes for all 5 major parties. No one else, including EKOS, was close.

Then again, anyone who would use the bandwidth yesterday that you did extolling the liquidity of our banks only to wake up to today's bailout can't be expected to demonstrate much objectivity.

What bailout?

[Jobu]

Nanos has Tory lead back up to 6.

http://www.nanosresearch.com/election/CPAC...er-10-2008E.pdf

[Vancouver King]

What was the final margin in 2006? Nanos again confirms some significant differences from 2006.

Quebec remains a Tory wasteland this campaign, their losses there will be a minimum 4 seats. Add that to the 3 they will lose in Nfld. and the party will be down about 7 ridings before it gets to the Ontario border.

Seems it all boils down to Ontario - who can argue it should be settled anywhere else.

[Jobu]

Here's the commentary:

"As of last night we have six point gap between the Conservatives and the Liberals (33% to 27%). Conservative support remains unchanged but there has been movement from the Grits to the NDP (who we now have at 22% nationally).

A few things to consider going into the final weekend of the campaign. First, if the Conservatives can pivot the focus of the election from themselves to Dion, it will serve them well. The reality is that as long as the Harper Tories drive support away from the Liberals (regardless of the final voter destination), it helps the Conservatives.

Because a Conservative majority win is unlikely at this time, the NDP will have a higher likelihood of holding on to their support because of less strategic voting.

For the Dion Liberals, they will likely battle to keep the focus on Harper to make him the ballot issue."