Friday, February 11, 2011

Bonus TWIA - Polls!

Two new polls to report in Alberta.

First up comes the latest from Janet Brown, who has earned a solid reputation for projecting Alberta elections:

PC 40%
WAP 33%
ALP 19%
NDP 8%
AP 0.1%

I feel badly for Brown on this one, since this poll was in field the week before Stelmach resigned. A 900 sample survey isn't cheap, and with the oil hitting the fan right afterwards, the poll loses a lot of its value. But, hey, at least we know the PCs were doing fine, even with Ed at the helm.

Luckily, Environics was in field both before and after all the craziness. Their conclusion? It didn't change a heck of a lot. They didn't notice an immediate shift in support one way or the other.

PC 38%
WAP 26%
ALP 22%
NDP 10%

Numbers like this have a tendency to get spun as "bad news" for all parties involved, so let me play the optimist for a change. The PCs are comfortably in front, likely in majority territory. The Alliance are a solid second despite most Albertans knowing little about them or their charismatic leader. The Liberals are down, but if they can get their standard 25-30%, they'll make major seat gains due to the right wing vote split. The NDP are holding their ground, and could make inroads against a weak Liberal Party.



  • Can you use these numbers for seat projections like you do federally?

    By Blogger Robert Vollman, at 4:32 p.m.  

  • RV - I'm tempted to, but I'm really hesitant about projecting out the next Alberta election, just because I think the Wildrose Alliance is such a new entity that there's not a great baseline to work off.

    Let's assume there are two Calgary ridings - they got 3% in one last time and 6% in the other.

    If we do a uniform swing, then you're basically giving them the same support in both ridings (say, 33% and 36%)...and nearly all ridings (since there probably wasn't much variation last election).

    The flip side is if you do a geometric swing, then you're saying that the one riding would have 40% and the other would have 20%...which is a big assumption to make.

    So it's a tough one to do. I've seen a few projections out there (and Brown has one in the poll I linked to), but I'm not convinced on how accurate they'd be.

    By Blogger calgarygrit, at 6:23 p.m.  

  • Good point, but isn't there historical data somewhere that gives some indication how voters respond to hot new parties?

    For instance, the federal Reform party back in 1993? Compare it to 1988, and the polling numbers - would that help?

    By Blogger Robert Vollman, at 10:37 p.m.  

  • Yeah, I was thinking Reform '93, or even ADQ 2007 (?) would be good case studies.

    By Blogger calgarygrit, at 9:31 a.m.  

Post a Comment

<< Home