Today's SES poll reports a statistical tie between the Liberals and Conservatives, suggesting immediate and substantial damage from the income trust investigation. The 34% for the Conservatives probably translates into about 43/44% in Canada excluding Quebec.
Good news for Stephen Harper with one catch. Now that he may be about to be perceived as the front runner, the public will want to scrutinize him more closely.
On a loosely related note, the Conservative tactic for dealing with anticipated negative Liberal ads that one might now anticipate is to run this bizarre ad of their own warning us against the dastardly Grits.
Subscribe To
Posts
All Comments
Friday, December 30, 2005
More Harris Legacy
This story in the Thursday, December 29 Globe about charges against Ontario Power Generation had a paragraph that jumped out at me. The excerpt is longer to put it in context but I have put the relevant paragraph in italics:
“Along with two employees, OPG is charged with two counts of criminal negligence causing death and seven counts of criminal negligence causing bodily harm.
The charges were laid in connection with a June, 2002, incident in which a rarely used gate at the Barrett Chute hydroelectric dam on the Madawaska River was opened to drain off excess water, engulfing a group of sunbathers in a surge of water that swept them over a 10-metre-high cliff and onto rocks below.
Homicide charges against a publicly owned utility appear to have no precedent in Canada, said Ontario Provincial Police Staff Inspector Ian Grant, who headed the exhaustive two-year criminal probe into the incident, which he terms "a very difficult investigation."
Moreover, the trial, which will be by a judge alone, is sure to focus on the decision-making process at Toronto-based OPG.
Employees have said that after the Mike Harris government's decision to deregulate the electricity market -- a move that took effect just weeks before the tragedy -- local control over dams such as Barrett Chute was transferred to a computerized dispatch system at head office.”
Could this be Walkerton 2? What should it tell us about what to expect from the former President of the National Citizens’ Coalition who now thinks he should be Prime Minister, so he can promote, among other things, privatization and deregulation?
“Along with two employees, OPG is charged with two counts of criminal negligence causing death and seven counts of criminal negligence causing bodily harm.
The charges were laid in connection with a June, 2002, incident in which a rarely used gate at the Barrett Chute hydroelectric dam on the Madawaska River was opened to drain off excess water, engulfing a group of sunbathers in a surge of water that swept them over a 10-metre-high cliff and onto rocks below.
Homicide charges against a publicly owned utility appear to have no precedent in Canada, said Ontario Provincial Police Staff Inspector Ian Grant, who headed the exhaustive two-year criminal probe into the incident, which he terms "a very difficult investigation."
Moreover, the trial, which will be by a judge alone, is sure to focus on the decision-making process at Toronto-based OPG.
Employees have said that after the Mike Harris government's decision to deregulate the electricity market -- a move that took effect just weeks before the tragedy -- local control over dams such as Barrett Chute was transferred to a computerized dispatch system at head office.”
Could this be Walkerton 2? What should it tell us about what to expect from the former President of the National Citizens’ Coalition who now thinks he should be Prime Minister, so he can promote, among other things, privatization and deregulation?
Thursday, December 29, 2005
Mid campaign polls and accidents
An Environics poll released on Christmas Eve noted, “…the federal election campaign has yet to have any noticeable effect on voter support.” Using that statement as my cue (and I do agree) I averaged all the polls taken since the beginning of the campaign including the daily SES (except for today’s release) and the Globe’s Strategic Counsel. This calculation does not take overall movement within the campaign into account, but then there has not been much, at least in aggregate.
I would say the ups and downs in the polls suggest there is some non-sampling error going on. For example, estimates of Liberal strength in Ontario have ranged between 37% and 47% although most readings are in the forties. In B.C. the range has been between 26% and 42% with most reading in the thirties.
Applying my seat forecaster to the overall average produces the following results: Liberals – 131, Conservatives – 87, NDP – 23, BQ – 67. For full details of the poll averages and seat numbers see here.
However, I began writing this post before the news of the RCMP investigation of the so-called Income Trust leak. I think this story is potentially quite damaging to the Liberals (even though I believe in the end the investigation won’t produce results). It plays into, and reinforces the Liberal corruption narrative promoted by the opposition. I think it more likely to be helpful to the Conservatives than the NDP (despite their role in producing the investigation) as the Tories have run with the corruption issue more strongly.
This is a campaign in search of a galvanizing moment. Is this it? Perhaps, but it is too soon to tell. It also comes in a week where the Liberals were afflicted by other ‘accidents’ in what has seemed an accident-prone campaign.
I would say the ups and downs in the polls suggest there is some non-sampling error going on. For example, estimates of Liberal strength in Ontario have ranged between 37% and 47% although most readings are in the forties. In B.C. the range has been between 26% and 42% with most reading in the thirties.
Applying my seat forecaster to the overall average produces the following results: Liberals – 131, Conservatives – 87, NDP – 23, BQ – 67. For full details of the poll averages and seat numbers see here.
However, I began writing this post before the news of the RCMP investigation of the so-called Income Trust leak. I think this story is potentially quite damaging to the Liberals (even though I believe in the end the investigation won’t produce results). It plays into, and reinforces the Liberal corruption narrative promoted by the opposition. I think it more likely to be helpful to the Conservatives than the NDP (despite their role in producing the investigation) as the Tories have run with the corruption issue more strongly.
This is a campaign in search of a galvanizing moment. Is this it? Perhaps, but it is too soon to tell. It also comes in a week where the Liberals were afflicted by other ‘accidents’ in what has seemed an accident-prone campaign.
Tuesday, December 20, 2005
Decima Polls
A new Decima poll out today has the Liberals up 10 points on the Conservatives 35% to 25% with the NDP at 21% and the BQ at 13%.
It is one of a series of polls conducted via the internet. I think that net polling is where we are headed in the future and is going to be increasingly important, but one should note that these election polls I think are best described as an experiment. For example, see this description of the methodology on this news release from Decima. Note the section I underlined below, which is an important qualification.
(Decima has launched an innovative on-line voter tracking study in collaboration with the Carleton University School of Journalism and Communication. The data reported here was gathered over the period November 26 through December 1st, from 4337 survey participants, drawn from Decima’s proprietary eVox online panel. For a comparable random probability sample, the findings would be considered accurate to within 1.49 percentage points, 19 times out of 20. Decima plans to re-interview those who agree to be part of a special election-tracking panel each week leading up to the election.)
I do think these results are worth watching, and Decima has apparently pre-selected this panel in the traditional way but the approach remains quite new so some caution is in order.
It is one of a series of polls conducted via the internet. I think that net polling is where we are headed in the future and is going to be increasingly important, but one should note that these election polls I think are best described as an experiment. For example, see this description of the methodology on this news release from Decima. Note the section I underlined below, which is an important qualification.
(Decima has launched an innovative on-line voter tracking study in collaboration with the Carleton University School of Journalism and Communication. The data reported here was gathered over the period November 26 through December 1st, from 4337 survey participants, drawn from Decima’s proprietary eVox online panel. For a comparable random probability sample, the findings would be considered accurate to within 1.49 percentage points, 19 times out of 20. Decima plans to re-interview those who agree to be part of a special election-tracking panel each week leading up to the election.)
I do think these results are worth watching, and Decima has apparently pre-selected this panel in the traditional way but the approach remains quite new so some caution is in order.
Monday, December 19, 2005
The Static Campaign
This campaign has a static feel too it in the sense that nothing seems to be happening. The first debates are over but no doubt the leaders are holding their best lines for January.
These polls conducted by Decima and Strategic Counsel (on page 28) tell us that Layton benefited the most from the debates. I reject the bare bones questions about who won as misleading given that they appear simply to mimic underlying voter preferences.
The finding by Strategic Counsel (on page 30) that 95% of their poll’s respondents found nothing in the debates to make them change their votes confirms to me the sluggish character of this campaign.
It appears that we can’t expect to see much dynamism before January 2, if then.
These polls conducted by Decima and Strategic Counsel (on page 28) tell us that Layton benefited the most from the debates. I reject the bare bones questions about who won as misleading given that they appear simply to mimic underlying voter preferences.
The finding by Strategic Counsel (on page 30) that 95% of their poll’s respondents found nothing in the debates to make them change their votes confirms to me the sluggish character of this campaign.
It appears that we can’t expect to see much dynamism before January 2, if then.
Thursday, December 08, 2005
One more observation
The one constant in most polls this week is a weakening of Bloc support in Quebec. My intuition about this is that Quebec will split largely on federalist/sovereignist lines and produce results similar to those of 2004. Martin's call for referendum like voting is certainly in his self interest and not therefore surprising but I suspect it was unnecessary.
Contradictory Messages in Polls
Since most polls have reported the Liberals to be leading, there is a general impression of not much happening in this campaign. It has had a low key flavour.
However, the polls are giving quite contradictory messages. I think this deserves more attention. An Ipsos-Reid poll completed last week did have the Liberals ahead but only by two points, 33% to 31% over the Conservatives with the NDP at 17%. Once the regional numbers are crunched through my forecasting model we find a Conservative minority government with Mr. Harper’s party winning 115 seats to the Liberals’ 97 and the NDP’s 29.
Contrast that with today’s SES poll. It has the Liberals ahead overall 40% to 26% with the NDP at 18%. The data translate into a Liberal majority government of 168 seats with the Conservatives’ 57 seats apparently putting them in third place behind the Bloc’s 59 seats, with the NDP at 24 seats.
The Globe’s Strategic Counsel numbers are in between.
The difference cannot be explained by the half a week lag between the two surveys. I continue to believe current preferences are quite weak, and what we are really seeing is differences in methodology. I think it likely that one pollster or the other has got it wrong (or both), but there is no immediate election to tell us which one with any certainty.
Personally, the numbers I find hardest to believe in the two surveys are the SES numbers for Western Canada, which report a massive swing away from the Conservatives to the Liberals and the NDP. In 2004 the results in the West were Liberals 26.8%, Conservatives 45.5% and the NDP 20.4%. Today’s SES poll the West region preferences are given as Liberal – 35%, Conservative – 36% and NDP – 24%, quite a shift.
The polls should converge by the end of the campaign.
Addendum: Paul Wells seems to have spotted this emerging problem. A new Léger Marketing survey appears to support SES. However, Léger has uniformly delivered numbers very favourable to the Liberals so I don't find their results all that surprising. Their western numbers show a significantly smaller shift to the Liberals than SES.
However, the polls are giving quite contradictory messages. I think this deserves more attention. An Ipsos-Reid poll completed last week did have the Liberals ahead but only by two points, 33% to 31% over the Conservatives with the NDP at 17%. Once the regional numbers are crunched through my forecasting model we find a Conservative minority government with Mr. Harper’s party winning 115 seats to the Liberals’ 97 and the NDP’s 29.
Contrast that with today’s SES poll. It has the Liberals ahead overall 40% to 26% with the NDP at 18%. The data translate into a Liberal majority government of 168 seats with the Conservatives’ 57 seats apparently putting them in third place behind the Bloc’s 59 seats, with the NDP at 24 seats.
The Globe’s Strategic Counsel numbers are in between.
The difference cannot be explained by the half a week lag between the two surveys. I continue to believe current preferences are quite weak, and what we are really seeing is differences in methodology. I think it likely that one pollster or the other has got it wrong (or both), but there is no immediate election to tell us which one with any certainty.
Personally, the numbers I find hardest to believe in the two surveys are the SES numbers for Western Canada, which report a massive swing away from the Conservatives to the Liberals and the NDP. In 2004 the results in the West were Liberals 26.8%, Conservatives 45.5% and the NDP 20.4%. Today’s SES poll the West region preferences are given as Liberal – 35%, Conservative – 36% and NDP – 24%, quite a shift.
The polls should converge by the end of the campaign.
Addendum: Paul Wells seems to have spotted this emerging problem. A new Léger Marketing survey appears to support SES. However, Léger has uniformly delivered numbers very favourable to the Liberals so I don't find their results all that surprising. Their western numbers show a significantly smaller shift to the Liberals than SES.
Tuesday, December 06, 2005
Polls - Much Ado About Nothing
There are two tracking polls in this election, one from SES for CPAC, the other from Strategic Counsel for the Globe/CTV. Both show the Liberals ahead but little apparent impact from the campaign. The commentary from the pollsters strikes me as being for the benefit of making news for their respective media, and overstates the significance of what is really happneing. The data regionally do show significant variations between the two pollsters with SES showing more favourable numbers for the Liberals.
In any event current preferences are weak regardless of distribution. I don't think the public is yet focused so the numbers are a benchmark but little else at this stage. I continue to think the television debates this time have the potential to have a great impact. They begin December 15.
In any event current preferences are weak regardless of distribution. I don't think the public is yet focused so the numbers are a benchmark but little else at this stage. I continue to think the television debates this time have the potential to have a great impact. They begin December 15.
Friday, December 02, 2005
Harper's GST Promise
Just a quick note to say that Harper has done well with his gst promise and the Conservative campaign overall appears to be reflecting the fact they have now had 18 months to get themselves better organized than last time.
However, most polling I have seen since the federal government went into surplus suggests most voters put a higher priority on social spending (health in particular but also education, etc.) . See these recent CRIC poll numbers for example. The media I think should pay more attention to the appeal of this option not just income vs gst tax cuts.
In any event it is far from clear yet that the election will end up being about taxes.
However, most polling I have seen since the federal government went into surplus suggests most voters put a higher priority on social spending (health in particular but also education, etc.) . See these recent CRIC poll numbers for example. The media I think should pay more attention to the appeal of this option not just income vs gst tax cuts.
In any event it is far from clear yet that the election will end up being about taxes.
Subscribe to:
Posts (Atom)