It's amazing how the media reports polling from PPP without mentioning that it is a Democrat polling firm. We have previously pointed out how they over sample Democrats. Well, they were at it again this week in California where they came out with a poll that had Boxer over Fiorina 52-43, and Brown over Whitman 53-42. The problem is that their sample was more unfavorable to Republicans than the 2006 turnout. 2006 was the worst year for the GOP since Watergate and there were no significant Republican candidates on the ballot to turn out the Republican base. Yet, we are to believe that not only will turnout be at least as good as 2004, it will be worse than 2006.
The 2006 exit poll data from California showed the following partisan turnout: 41% D, 35% R, 25% I. The PPP poll shows a partisan breakdown of 47% D, 34% R, 19% I. Also, in 2006, women constituted 51% of the electorate, while they represent 53% in the PPP survey. Does anyone really believe that turnout models will be worse than in 2006?
Also, the latest Suffolk University poll is even worse. They also have Boxer leading 52-43, but their survey sample includes 45% D, 31% R, and 25% I. It is important to discount any Suffolk poll for the rest of the election season because of their ridiculously callous way of screening for likely voters. Sean Trende, at RCP has a good write-up on this point.
No comments:
Post a Comment