The infinite edition Polling Thread: Volume 5

98 posts / 0 new
Last post
Sean in Ottawa

Then he should pretend he is talking to kids when he is speaking to the media--

In many respects they haven't grown up either.

Sean in Ottawa

Good to hear though-- I think he is going to have an opportunity in the next election - it is critical that he does not lose that chance

remind remind's picture

He looks great in a blue shirt and a ballcap, say no more.... ;)

Sean in Ottawa

Shhhh the Blue Jays will draft him and we can't pay what they can...

And you never have to win if you work for a Toronto sports team.

ottawaobserver

Sean in Ottawa wrote:

Then he should pretend he is talking to kids when he is speaking to the media--

In many respects they haven't grown up either.

Haha.  They're not going to pay attention for two hours, though.  :-)

KenS

ottawaobserver wrote:
Plus, I should add that young people I know say their peers think Layton was cool back when us older folks thought he was a bit too much of a self-promoter. 

Interesting, because I found this way back.

When Jack was first talking of running for the leadership, I went to his first Nova Scotia meet and greet. All I knew about him was his basic bio. Since there weren't any more than 20 people we could all get around a table. Much better for give and take than the usual schmoozing around stuff.

I stayed late so I could watch Jack, and mull. I don't know if I ever said a word. I concluded:

[1] He definitely had my support.

[2] I was pretty sure he was someone I wouldnt like as a personal acquaintance.

Couple days later I saw the young people who had been there- at least one of whom I know to be both observant and demanding. I expected them to be as impressed as they were. But they surprised me by using a lot of words like sincere, and genuine.

Conversely- my family that are my age, who were virtualy all NDP inclined before that became the norm in Nova Scotia, mostly think Jack is a phony. It makes for an intersting discussion theme when I point out that is not the way most people respond to him, because I know what they are talking about and on a visceral/experiential level I'm a long way from really understanding Jack's basic appeal.  

NorthReport

 

Hébert: Conservatives losing PR war
Opposition makes gains as Tory government's credibility crumbles

To sum up: the opposition raised its game; the government did not, and enough Canadians took notice to potentially make this end-of-session debate a watershed moment in the life of the minority Parliament.

http://www.thestar.com/news/canada/afghanmission/article/737539--hebert-conservatives-losing-pr-war

melovesproles

Quote:
I happened to be in the area, so I went to see Layton talk to students at the University of Ottawa the first week of September, as part of his annual campus tour.  You would have been really impressed Sean.  Not a canned answer in the two hours, and in fact he spent very little time with the opening statement, spending most of the two hours taking questions and (more importantly) listening to what the young folks had to say.  He was completely spontaneous, very thoughtful, pitched his comments just right (not too obtuse, but not dumbed down), and demonstrated his command of a wide-ranging array of topics while making the links between then.  Not a scripted message anywhere to be found.

Layton definately has that ability, I've been impressed by him a number of times, which is why it's frustrating that this isn't the way he usually deals with the media.  If you're attending political talks at Universities or hunting down obscure radio interviews then you're probably acquainted with that side of Layton but that's mostly people who are on the political junkie side of the divide.  I think Sean is spot on with his suggestions for a new approach to party communications.

NorthReport

So if the Cons had dropped to only  33%  on the last day of the most recent EKOS poll, the NDP must be doing very well, probably in the  20%  range. 

 

http://www.theglobeandmail.com/blogs/bureau-blog/stephen-harpers-salad-days-are-done/article1395696/

NorthReport

This is one scary poll. The Cons have never hit 50% support in Manitoba before. 

Federal polling for Manitoba

Cons - 50%

NDP - 22%

Libs - 21% 

Conservatives hit magic 50

 

 

Half of Manitoba voters backing federal Tories

 

 

The inability of Liberal Leader Michael Ignatieff to make headway with voters is apparent in the poll. The Liberals had climbed up to 26 per cent among Manitoba voters last June. But in the last six months, the Liberals have lost almost all the ground they gained since the last election and the disaster of former Liberal leader Stéphane Dion and his ill-advised Green Shift platform.

 

http://www.winnipegfreepress.com/local/conservatives-hit-magic-50-792020...

Stockholm

I'm not sure what the big deal is. The Tories took 49% of the vote in Manitoba in the last election and this poll has them at 50% - so what? I was not aware that there was anything "magical" about getting 50% of the vote in one small province after having had 49% in the last election.  If an election were held and these were the province wide popular vote numbers - not a single seat in Manitiba would change hands.

NorthReport
madmax

That poll above is a joke. JOKE. 

It looks like polsters are going the way of the dodo bird.

ottawaobserver

Angus Reid is a joke? That's what "Vision Critical" is now. They have a consistent methodology and a record of results to compare with. I would have characterized their pools as among the least laughable in general.

What made you call that poll a joke, madmax?

There were results reported for the Bloc (11%) and Greens too (6%), by the way, along with some very interesting findings about preferances for forms of cooperation between the Liberals and NDP (the most popular of the 3 approaches is the post-election coalition).

NorthReport

Is that an Angus Reid Strategies poll?

If it is is, and I'm not sure it is, it's an unusual way of releasing poll results.

Stockholm

surprise! surprise! Tory supporters don't like the idea of the Liberals and NDP working together and/or forming a coalition. I guess that must because....such a thing would inevitably lead to the Tories losing power.

Stockholm

The one thing that does make sense is the Greens at 6% - though that is still probably an overestimate.

I don't find it that strange that the Liberals could get a bit of a "dead cat bounce' from the last survey. The Tories have had a bad week or so - and for most people the Liberals are still the kneejerk default option when they get disgruntled with the Tories.

The most notable thing that some polls are showing right now is that Tory support has been eroding bit by bit. We will need an election campaign to happen to know where everything shakes out between the Liberals and NDP etc...but the more that Tories numbers get suppressed the more it opens possibiliies of an election and a new post-election non-Conservative government.

Centrist

WTF? The Libs up 6%? Nothing makes logical sense here.

Are we sure that ARS didn't poll on Mars?!

BTW, has ARS now changed its name to Vision Critical? Silly name IMHO.

Sean in Ottawa

Yes Vision Critical is indeed Angus Reid (I just spoke to them on Friday).

It would surprise me if the Liberals were not up considering the detainee issue-- many who support the war do not like what this government is doing on that file or the attack on the diplomatic service. The story also underlines an ongoing problem of lack of transparency and hostility to criticism on the part of this government. It would have been nice to see the NDP higher but everybody is about where we would expect them to be. I will say including the Greens given Copenhagen.

Sean in Ottawa

 

Note this section of the analysis:

"Approval

Stephen Harper’s approval rating is 32 per cent (-2), while NDP leader Jack Layton gained five points since mid-November and is at 29 per cent. Liberal leader Michael Ignatieff gained three points (15%), but remains well behind his two main rivals.

Momentum

Ignatieff’s momentum score, while better than last month’s, is still the worst of all three party leaders at -33 (7% of Canadians report an improvement in their opinion of the Liberal leader over the past month, while 40% say their views have worsened). Layton improved from -10 in mid-November to -6 this month. Harper had the toughest month, going from -9 to -24."

Stockholm

Sean in Ottawa wrote:

....everybody is about where we would expect them to be. I will say including the Greens given Copenhagen.

Yes, isn't it interesting that a poll in the midst of the Copenhagen conference when environmental issues are dominating the news for the first time in over a year - Green support promptly collapses from 10% to 6%.

Debater

Centrist wrote:

WTF? The Libs up 6%? Nothing makes logical sense here.

Why doesn't it make sense that the Liberals might start going up again after having bottomed out?  The last few weeks have been bad for the Conservatives.

Popperian

The notion that polls which show  Green Party support higher than the last elecition are not "true" measures of Green support  shows a (commmon) misunderstanding of the nature of polling between elections. The parameter which we are interested in estimating is the incidence of all eligble voters who currently prefer the Green Party . To compare this to the last election and conclude that  the lower figures must be "right" confuses the issues of predicting some future hypothetical election with the problem of modelling current support levels amongst the overall population of eligible voters.
Clearly the Green Party support is higher amongst eligible voters than it has been amongst the smaller population of voters who will actually vote. This problem of the gap between eligible and likely voters becomes more meaningful as we approach an actual election . It may , in fact make sense at that point to report two separate estimates (likely and eligble voters). At this point, however, it makes little sense to do so now and it is ill advised to discount those who  select the Green Party when given a choice in a poll  (even knowing that a large numbe of them will not actually vote) By not prompting with GP one will achieve a lower estimate but this will bias the estimate of how all eligible voters are choosing . A correction for likely voters will make the proper adjustment as an election looms . In the interim I have no doubt that the higher estimates are a better measure of the overall popularity of the Green Party within the overall population of eligible voters.
We are not interested in predicting a hypothetical future election now and it is the lower estimates  which are indeed the biased indicators of the current intentions of the entire eligible population. It is , however, important to note the caveat that current GP supporters are hsitorically much less likely to actually vote than supporters of other parties. It remains to be seen if the attrition from eligible to actual voters will be of the same magnitude in the next election as it was in the past one

madmax

Popperian wrote:
 By not prompting with GP one will achieve a lower estimate but this will bias the estimate of how all eligible voters are choosing .
Then prompt for all parties including the Rhino Party. Watch the Rhino numbers go up in the off season and down come Eday or maybe not.

If a polster isn't close on EDAY then their polls have no value and lose their credibility. It is so bad, that Ekos is making excuses for their inflated numbers. 

If you are poll at 11% to 15% before the vote and end up with 3% and 4% at Eday then the pols are WRONG and can only be taken with a grain of salt.

Polster can and do make corrections with the data.  

Did you look at the last polls western results?  ZERO , ZERO votes for the NDP and the Green Party at 20%.  Based on how many votes? 7 VOTES out of a Cross Canada Sample of 500. 

Now anyone in Manitoba would know the NDP are going to be in the 20% range to as high as 47% if Provincial. Even still the National polling numbers are often within the margin of error for the NDP, CPC and LPC.

Does the polling company have to create a margin of error of 200% for one political party?

 

NorthReport

At least May won't be in the debates this time as there are no Green MPs unless of course Stephane Dion jumps ship. Laughing

KenS

Popperian wrote:

Clearly the Green Party support is higher amongst eligible voters than it has been amongst the smaller population of voters who will actually vote. This problem of the gap between eligible and likely voters becomes more meaningful as we approach an actual election . .... Etc.

Problem is that this is only one of the reasons for the consistent considerable variance beween polled and voting support. The other being that a much higher proportion of those who say they will vote for the GPC actually do vote, but for another party.

bekayne

madmax wrote:

Popperian wrote:
 By not prompting with GP one will achieve a lower estimate but this will bias the estimate of how all eligible voters are choosing .
Then prompt for all parties including the Rhino Party. Watch the Rhino numbers go up in the off season and down come Eday or maybe not.

The difference is that the Greens run close to full slate while the other minor parties struggle to get 50 candidates

 

Debater

NorthReport wrote:

At least May won't be in the debates this time as there are no Green MPs unless of course Stephane Dion jumps ship. Laughing

Which obviously isn't going to happen since that story died out weeks ago and wasn't true anyway.

May will have to try and convince someone else to commit political suicide by running for party that can't win a seat.  The moment you cross the floor and become a Green, your political career is over.  Just ask Blair Wilson.

NorthReport

Cons beginning to sink?

 

Harris/Decima

C - 34%

L - 28%

N - 14%

http://www.theguardian.pe.ca/index.cfm?sid=311312&sc=117

NorthReport

---

Stockholm

Debater wrote:

The moment you cross the floor and become a Green, your political career is over.  Just ask Blair Wilson.

Blair Wilson's political career was over regardless. He had been charged with corruption and expelled from the Liberal caucus. I guess he figured that while his chances of winning as a Green were low - if he had run as an independent his chances would have been non-existent!

I suppose that if May really thinks that getting any warm body to be a GINO (GReen in name only) for the purposes of making the case for her to be in the debate is a high priority - there will always be someone who has been expelled or lost a nomination battle who wants 15 minutes of fame as a Green MP before being crushed. maybe she can go after Ruby Dhalla!

Stockholm

"So let's stop fantisizing about those large numbers of Green voters who change their minds or stay home-- they probably don't exist and the pollsters have not found a way to avoid over-representing enthusiastic but small numbers of Green voters in opinion polls."

I have to disagree - I just don't believe that there is this vast movement of Green supporters who hear the phone ring and see on the call display that its Ipsos or whoever and exclaim "OMG, a polling company is calling me - I've got to answer right away so i can say I'd vote Green - jut so that they get better polling numbers". Keep in mind that for every poll on political preference - there are dozens on consumer products etc... those mythical Greens who go out of their way to participate in every poll under the sun just to drive up their party's polling numbers would have to endure one hell of a lot of polls about how to get the yellows out of floor wax!

I'm still impressed by how when Nanos does a poll and does not prompt for any parties at all - Green support collapses into low single digits - if Green supporters were really so gung ho about being polled you would think that they would not need to be reminded that their party existed by hearing it read to them in a list of parties - they would just spontaeously say GREEN when asked "who are you going to vote for?" But they don't - and the fact that evidently half of the people who say they would vote Green in a poll only do so when the party name is prompted - tells me that these are very weakly motivated people with low political literacy.

Sean in Ottawa

KenS wrote:

Popperian wrote:

Clearly the Green Party support is higher amongst eligible voters than it has been amongst the smaller population of voters who will actually vote. This problem of the gap between eligible and likely voters becomes more meaningful as we approach an actual election . .... Etc.

Problem is that this is only one of the reasons for the consistent considerable variance beween polled and voting support. The other being that a much higher proportion of those who say they will vote for the GPC actually do vote, but for another party.

I think you are both wrong.

You are not alone. Polling companies won't admit that their numbers are not accurate to begin with so they come up with explanations of non-voters being polled and people changing their minds. Here are the false assumptions you and several others are making:

1) The number who are polled reflect the voting intention of the public.

2) When the numbers are different we start to make assumptions that either the people who say they will vote Green don't vote or do vote for other parties come election time.

3) People are more likely to be polled than to vote (this is the cornerstone of the misconception).

All of the above assumptions are wrong. Here are some reality checks:

1) The number of people willing to be polled is minuscule compared to the number willing to vote and the gap is getting wider (the number willing to do surveys is decreasing even faster than the number willing to go out and vote-- survey fatigue is part of it). Pollsters are now finding it increasingly difficult to produce statistically relevant data. When there is a difference they want to explain that somehow people changed from when they did the "accurate" poll to when they voted rather than allow that the poll was not accurate to start with.

2) There is no evidence that Green supporters are any less likely to vote than those of any other party (even among youth). I would argue that they are perhaps more likely to vote-- AND more likely to participate in polls and that is where the explanation for the difference between polls and elections comes from.

If one party's support is more committed than another's, that party will be over represented in polls not the reverse -- there are many reasons including voter anger, idealism etc. for people to be motivated to tell a pollster how they feel.

Remember the so-called ballot bonus for the Liberals- that's the flip side? The Liberals have a lot of support that often is not motivated to make a point by talking to a pollster. NDP and Con voters are both more likely to be motivated to do a poll. Green voters fall over themselves to be polled- they are a very small group, many are activists wanting to make a point. Because they are always willing to drop everything to be polled they are over represented when other parties supporters won't take the pollster's call.

What I am saying may be controversial but it is this: Green voters don't change their minds or decide not to vote-- it is just that there are fewer of them than we think because of self-selection in polling.These are the true believers who will vote for a party that has nil chance of winning even a handful of seats.

If you made everyone vote, the Green vote would go down rather than up--

An election is a better more reliable measure of people's opinions poll than an opinion poll -- Don't let the pollsters convince you otherwise.

So instead of saying clearly the Green party support is higher among eligible voters (when there is no evidence for that) it would be accurate to say the Green party support is higher among those who are willing to be polled than those who are willing to go and vote. And the much bigger number is those willing to vote.

So let's stop fantasizing about those large numbers of Green voters who change their minds or stay home-- they probably don't exist and the pollsters have not found a way to avoid over-representing enthusiastic but small numbers of Green voters in opinion polls.

Sean in Ottawa

I used to be involved in polling and political polls are very self-selective.

People do not realize how that works.

As a rule pollsters front end the questions on politics because easily 10 times the number will do the poll if it is about politics compared to laundry soap. In order to save money the pollster writes a short intro blurb that includes the nature of the poll in the first seconds to try to get those who would hang up on other polls not to on this one. The interviewers even try to bust refusals by asking people if they want to be counted-- I know, I trained many to do just that.

And there are not legions of green supporters-- that's my point but those inclined to vote Green get excited by doing a poll-- I have heard thousands of such calls. The only party without enthusiasm is the Liberal party which is chronically under-represented in polls. But nobody turns on like Green supporters. I have also heard the interviewers bust refusals and then I hear the results. The ones who are reluctant but can be convinced are for the other parties-- the ones anxious to participate as soon as they know it is a political poll-- those are overwhelmingly Green supporters unless there is a current issue where voters are fired up (like when a government is really about to be defeated becasue people are angry and then everybody wants to participate). People get more interested the close the election gets. You will notice that Green support goes down not just at election time but as we get close -- that's because the wider population gets interested.

ottawaobserver

Popperian wrote:

The notion that polls which show  Green Party support higher than the last elecition are not "true" measures of Green support  shows a (commmon) misunderstanding of the nature of polling between elections. The parameter which we are interested in estimating is the incidence of all eligble voters who currently prefer the Green Party . To compare this to the last election and conclude that  the lower figures must be "right" confuses the issues of predicting some future hypothetical election with the problem of modelling current support levels amongst the overall population of eligible voters.

Welcome to Babble, Popperian.  I'm guessing you're associated with Ekos and/or the Carleton School of Journalism, since you're defending their methodology with the pronoun "we".

Popperian wrote:

Clearly the Green Party support is higher amongst eligible voters than it has been amongst the smaller population of voters who will actually vote. This problem of the gap between eligible and likely voters becomes more meaningful as we approach an actual election . It may , in fact make sense at that point to report two separate estimates (likely and eligble voters). At this point, however, it makes little sense to do so now and it is ill advised to discount those who  select the Green Party when given a choice in a poll  (even knowing that a large numbe of them will not actually vote) By not prompting with GP one will achieve a lower estimate but this will bias the estimate of how all eligible voters are choosing . A correction for likely voters will make the proper adjustment as an election looms . In the interim I have no doubt that the higher estimates are a better measure of the overall popularity of the Green Party within the overall population of eligible voters.

If you're trying to estimate the incidence of eligible voters who currently prefer the Green Party, and you assert that your methodology doesn't artificially elevate that estimate, then how do you account for the very large gaps between the findings of polling companies which prompt with the parties' names, and those who don't?  Are you making a claim that a methodology which prompts gives a more accurate estimate?

Popperian wrote:

We are not interested in predicting a hypothetical future election now and it is the lower estimates  which are indeed the biased indicators of the current intentions of the entire eligible population. It is , however, important to note the caveat that current GP supporters are hsitorically much less likely to actually vote than supporters of other parties. It remains to be seen if the attrition from eligible to actual voters will be of the same magnitude in the next election as it was in the past one

In fact you *are* claiming the other methodology is less accurate: "it is the lower estimates which are indeed the biased indicators of the current intentions of the entire eligible population".

But, do people really need to be prompted in order to remember who's popular with them?  If so, is that a true measure of a party's "overall popularity"?  "You have no doubt they are", but they sure didn't predict actual voting behaviour during the by-elections, did they.

Meanwhile, given you have those caveats on what the results mean (they represent the "popularity" of the party with "eligible" not necessarily "likely" voters), you're in effect admitting that doing seats projections based on those numbers is not valid now, and won't be until closer to the next election when you decide it's worthy "correcting for likely voters".  Right?  Meaning, maybe you should stop doing those seat projections now.

The hypothesis of "attrition from eligible to actual" is not really capable of being tested, is it?  Not if you say that measuring eligible vs. likely "makes little sense at this point".  I assume you didn't poll around the by-elections in order to test this hypothesis.  Is there any evidence from the Canadian Election Study to support it?  What makes you so sure it isn't just a socially-acceptable response instead of admitting they won't vote?  Is there any methodological way to distinguish between the two hypotheses as the better explanations?  Are you so sure those same people don't wind up voting another way on election day rather than stay home?

In fact, colour me sceptical, but it sounds like you're a lot more concerned about accurately measuring the potential Green vote than accurately predicting the actual Liberal and NDP vote.

Anyways, thanks at least for showing up here to discuss the issue.  I see you weren't quite brave enough to approve my two comments on the blogpost at the Ekospolitics.com website.  No point "prompting" for commentary there if you're not going to publish it, is there.

NorthReport

Even though I don't always like the results, and Quebec polling is done well by CROP & Leger, Nanos seems to have the best methodology and usually comes closest to the overall final results. 

NorthReport

EKOS Every Other Weekly: Between the three of them, they're up nearly a full percentage point! Tongue out

 

 

http://www.cbc.ca/politics/insidepolitics/2009/12/ekos-every-other-weekl...

Debater

So according to that poll, the BQ is the main party that has dropped.

West Coast Greeny

NorthReport wrote:

Cons beginning to sink?

 

Harris/Decima

C - 34%

L - 28%

N - 14%

http://www.theguardian.pe.ca/index.cfm?sid=311312&sc=117

G - 13%

Just sayin'.

Stockholm

...and and Angus Reid says the Greens are at 6% and falling - and given the 2%, 3%, 2% and 4% showings in the recent byelections held in what should have been prime Green territory - I tend to think the Angus Reid poll is about right - though probably still an overestimate.

Debater

The EKOS poll provides some good news to Ignatieff for a change - he is solidly in second place in Quebec and is also a strong second in Ontario.

KenS

We'll see for how long.

Debater

Indeed.

Ken Burch Ken Burch's picture

Since this is getting close to 100 posts, I've started a continuation thread:

http://rabble.ca/babble/canadian-politics/polling-thread-beyond-infinite...

Stockholm

Debater wrote:

The EKOS poll provides some good news to Ignatieff for a change - he is solidly in second place in Quebec and is also a strong second in Ontario.

Its pretty sad when Liberals start to hail being a "strong second" in Ontario as good news. In other words its "good news" that the Liberals are behind in Ontario by the same margin that caused them to lose scores of seats in the last election. I guess you could say that the 2005 election was good news for the Liberals since with 77 seats they were a "strong second" to the Tories!

Debater

Stockholm wrote:

Debater wrote:

The EKOS poll provides some good news to Ignatieff for a change - he is solidly in second place in Quebec and is also a strong second in Ontario.

Its pretty sad when Liberals start to hail being a "strong second" in Ontario as good news. In other words its "good news" that the Liberals are behind in Ontario by the same margin that caused them to lose scores of seats in the last election. I guess you could say that the 2005 election was good news for the Liberals since with 77 seats they were a "strong second" to the Tories!

I disagree with that assessment Stockholm, but I don't feel the need to argue the point beyond saying that compared to the numbers the Liberals had earlier this fall, to be close behind the Conservatives in Ontario and ahead of them in Quebec is pretty good.  It also puts them well ahead of the NDP in both provinces.

NorthReport

The Liberals will be crashing once again as we step into the new year. Unfortunately they just keep picking the wrong leadership candidates. First Martin, then Dion, and now Ignatieff. Harper's running circles around them all and it shows. How stupid can the Liberals be.

Pages