More thoughts on the poll
First, check out what Jeff Kouba has to say over at BvW on the new SurveyUSA poll. His analysis seems sound and is worth reading. He discusses the impact of the most recent Wetterling ads and what role they may play in the new poll numbers.
Jeff also points out the crosstabs for this poll, and they tell an interesting story.
Look at the movement of the poll numbers across the three days of the poll:
Day 1, 10/6: Bachmann 50, Wetterling 42, Binkowski 6, Undecided 2
Day 2, 10/7: Bachmann 45, Wetterling 44, Binkowski 9, Undecided 1
Day 3, 10/8: Bachmann 46, Wetterling 45, Binkowski 7, Undecided 2
Wetterling's numbers jumped by 3 points, while Bachmann's fell by 5, across the three days of the survey. Now, because of the margin of error and the smaller sample size for each day, the movement might not mean anything. The degree to which Binkowski's numbers jumped around over the three days is a confounding factor for this theory. But one could reasonably infer that the overall trend, if there is one, is the Foley scandal having its impact. Remember, Patty's ad on the Foley scandal began running 10/3, and she delivered the national radio address on 10/7. During the past week there has been an enormous wave of media coverage for Wetterling, and Bachmann in interviews has seemed pouty and angry about the attention being given to Wetterling. It is conceivable that there would be a lag time between the ad and its effects.
A surprising and depressing story told in the crosstabs is the degree to which self-identified "Democrats" and "Liberals" are crossing over to support Binkowski, and even (shudder) Bachmann. Seriously-- 9 percent of liberals and 9 percent of Democrats support Binkowski?! 9% of liberals and 7% of Democrats support Michele Bachmann?
I would like to retract my earlier statement about this poll overstating support for Binkowski. I did some research and found that the Independence Party candidate in 2002 got around 7.5% of the vote, so I feel it is conceivable that Binkowski could receive a similar number. Some liberals could be attracted to his platform, for instance, because of his criticism of Israel's invasion of Lebanon, which was not popular among many liberals, but which both Bachmann and Wetterling expressed support for. I sincerely hope that, as the election gets closer, many of this group will switch their vote to Wetterling as it becomes apparent that John Binkowski cannot win and that the consequence of voting for him may be to elect Michele Bachmann.
But I am completely at a loss to account for the 9% of liberals that say they'll vote for Michele Bachmann, in my opinion one of the most reactionary conservative candidates ever to run for public office in this state. What could account for this? Is it the gay marriage issue? I know anti-gay-marriage liberals, though I don't think they are single-issue voters. Perhaps it is something else?
Finally, I would like to ask for your help. The SurveyUSA poll counts as "likely voters" an even 50-50 split between men and women. My research reveals that, in the counties that make up the Sixth, there is a very even demographic split between the genders, perhaps slightly tilted . But conventional wisdom is that women vote in greater numbers than men. An extensive Google search turned up no records of past Sixth District election records that included breakdowns by gender. If anyone is able to find them, please let me know. They would be very helpful to see if the poll's assumptions about the gender split are accurate.
If they are not, and more women end up voting than men, it could have a very beneficial effect for Patty Wetterling-- she leads among women by a 50-41 margin.