Tuesday, March 17, 2009

New Poll: Angus-Reid Strategies

Angus Reid Strategies has released a new poll today. It was taken between March 10 and March 11 and involved 1,002 people. Here is the national result:

Conservatives - 35%
Liberals - 31%
New Democrats - 16%
Bloc Quebecois - 10%
Greens - 7%

The Ontario result:

Liberals - 39%
Conservatives - 38%
New Democrats - 12%
Greens - 10%

The decline of the NDP in Ontario continues. They need to do something. I haven't seen the NDP in the news for weeks.

The Quebec result:

Bloc Quebecois - 38%
Liberals - 29%
Conservatives - 16%
New Democrats - 13%
Greens - 3%

The Conservatives continue to have trouble in Quebec, but it isn't the doom and gloom of two recent polls putting them at 10% and 13%. Nevertheless, it is clear the fight is now between the Liberals and the Bloc, but it is difficult to tell whether the Liberal vote is super-concentrated or not.

The poll also had some leadership questions. Stephen Harper was named the best PM by 26% of those polled, followed closely by Michael Ignatieff at 24%. Jack Layton was far behind with 10%. Harper was most popular in Alberta (50%), but way behind in Quebec (11%). That is extremely problematic for him. Ignatieff's best score was in the Prairies (28%) and Quebec (27%), and he did worst in Alberta (18%). Layton did best in Atlantic Canada (15%), but had a dismal 3% in the Prairies.

There are some bright spots for the NDP leader, however. He was named most trustworthy of the three (34%), and also was considered the one who most understood the problems Canadians had (39%). Harper was named the best on the economy (31%). In terms of which leader inspires confidence, which is probably the most important indicator, 30% said Ignatieff, 26% said Layton, and 25% said Harper.

The projection has changed because of this new poll. The Conservatives are down to 136 seats, losing one seat to the Liberals in Ontario, who are now at 102. The national popular vote has changed as follows:

Bloc Quebecois +0.1
New Democrats +0.0
Liberals -0.0
Greens -0.1
Conservatives -0.1

In Ontario, the NDP has dropped 0.3 but the Liberals have gained 0.1, putting them ahead of the Conservatives for the first time since I have begun making projections. In Quebec, the Conservatives and Liberals have traded 0.1 to the Liberal advantage, and the NDP has taken 0.2 from the Greens.

For the Conservatives, this is a decent poll since they are still ahead of the Liberals nationally and are in a statistical tie in Ontario. It is also somewhat better for them in Quebec, though still very worrisome. Their 33% in Atlantic Canada is strong, and they continue to dominate in British Columbia. The Liberals should be worried about their showing in that province, but can be happy to be ahead of the NDP in the prairies and in a dead heat in Ontario. The Quebec numbers continue to look good, though the 36% in the Atlantic is potentially trouble.

The NDP has some positive to take from this poll, as it has them performing better in British Columbia than they have recently and their 13% in Quebec and 29% in Atlantic Canada are better numbers than they have seen for some time. The 12% in Ontario, however, is critical.

The Bloc continues to be steady, though they would like to see the numbers back over 40% soon.


  1. I like your' blog, and come here to see what's been coming out lately, but I feel compelled to comment on one facet. Given that the different polling companies obviously employ different methodology, and have consistently different results from one another, oughtn't you try to compare only Ipsos to Ipsos, Angus reid to Angus Reid etc? If you compare apples to oranges, ( or perhaps, Macintosh to Golden delicious), then you will systematically overstate the volatility of the electorate in your seat projection model. I know this isn't a scientific compilation of data, but we all know that statistics are often mis-used because of their resemblance to real, hard science.

  2. I try to take that into account by giving different weights to the polling results. Based on how closely the firms predicted the 2008 electoral result, I've assigned value to each firm's results. Angus-Reid, for example, is considered the best. I reduce each poll's weight every month, and the size of the poll is also a factor.

    Since I include all of the polls and get an average from them, I think what you're concerned about is taken into account. Since the electoral results in 2004, 2006, and 2008 are also in the model, there is some solid foundation to the data plugged into the system.


COMMENT MODERATION POLICY - Please be respectful when commenting. If choosing to remain anonymous, please sign your comment with some sort of pseudonym to avoid confusion. Please do not use any derogatory terms for fellow commenters, parties, or politicians. Inflammatory and overly partisan comments will not be posted. PLEASE KEEP DISCUSSION ON TOPIC.

Note: Only a member of this blog may post a comment.