Every pollster that Nate Silver analyzed, except Mason-Dixon, had a Republican bias. Rasmussen was obviously the worst, and by a lot. I want to know two things. First, will Silver continue to use Rasmussen in his models. And, second, what kind of effect does it have to have pollsters giving a distorted picture of Republicans’ strength all year long? I would think that helps Republican fundraising and hurts Democratic fundraising. Should Democrats band together and start a polling firm that issues deliberately biased polls in their favor? Why not? Rasmussen is doing it. Shouldn’t we do the same thing?
Maybe Democrats should start dismissing polls the same way that Republicans dismiss the media (and reality)?
Boo:
Did you see Rasmussen’s reply to Nate’s questions? We have known Rasmussen was biased for a long time. He worked for Bush/Darth Cheney ’04 after all. Not to mention, he’s going on NRO’s latest Alaska(or where ever they are going this time) cruise.
No, I didn’t see their response. Do you have it?
I’ll have to see if I can find it, but Rasmussen basically said to piss off. That he doesn’t reply to bloggers or people beneath him.
Well, the results are there for all to see, as demonstrated by Nate, who’s got himself a pretty respectable perch at the Times.
http://fivethirtyeight.blogs.nytimes.com/2010/11/04/rasmussen-polls-were-biased-and-inaccurate-quinn
ipiac-surveyusa-performed-strongly/
So if Rasmussen doesn’t want to answer, fine, that’s not going to help their credibility with the public or with the profession.
But don’t the Republicans dismiss only the media that does not serve them and their interests? And then they cow, shame and demonize that small part of the media not serving GOP interests until they are so marginalized as to be viewed as irrelevant or they cave and acknowledge as legitimate the Republican viewpoint? Isn’t that kind of how the “America is a center-right country” meme gain such a foothold?
And the poll issue works hand in hand with the Republican biased media view. They are able to create a much more fertile environment for a self-fulfilling prophecy of GOP dominance. Seems like it goes right back to Rove’s, “We create our own reality and you simply study it” statement.
Apologies for this seriously OT post, but just saw this appear at TPM a few minutes ago.
Olbermann Suspended By MSNBC
No other details
I’ve read some reports that it may have to do with Olbermann donating to Dem candidates.
And who broke that story: Politico. They inquired through information act to get the info.
Hum…I wonder what little birdie twirped into their ears…
As long as news organizations are consistent in banning their talent from making contributions, I don’t care, but it seems like a joke to air nakedly partisan shows and then complain that contributions undermine the appearance of neutrality.
It apparently is due to contributions to Dems. The hypocrisy in the media world surrounding shit like this is just aggravating. You have an entire television network whose sole reason for existence is to spread the Republican talking points and agenda and that is just fine and dandy; no conflicts there. But another network, probably under fear of being called a friend of the DFH’s, deem it necessary to slap the hand of an on-air person because they contributed to the other party.
Just confirms, without a doubt, that IOKIYAR.
Once again, our Liberal Media in action. What a freaking joke.
I’m not surprised that most pollsters had a slight Republican bias in their samples this year. It’s tough to predict what voter turnout is going to be like, and in a non-presidential year, given that statistical models are based on historical performance, you’re going to see some folks overshoot the mark. I’m not exactly sure of Nate’s methodology in computing the “bias” in his table, but given the small number of polls involved for most of the polling firms, just eyeballing it any bias of less than a half percent in the voting models is negligible and even a full percentage point is probably not to shabby for model building. Especially since the relevant historical models for Democratic turnout they probably have data for include 1998, 1994, and, if we’re lucky, 1978. The models for the other midterm election years probably over-estimate Democratic turnout given the general voting trends of the country.
But a deviation of 4% across 100 polls? Unless I’m completely misunderstanding what Nate means by “bias” that’s impressively bad. I’m not even sure an incompetent pollster could get numbers that bad by accident. Which means that in the future anyone who is using an aggregate of polls should do what Kos has done this year – show two results. One without Rasmussen and one with. (I thought Kos’s use of the two aggregates was very educational, because you could often see just how much adding Rasmussen to the mix biased the narrative in the “poll of polls” type of analysis. And since I’m positive that Scott Rasmussen is not an idiot I’m sure that’s one of the reasons his polls were so obviously out of whack this year.)
Re :Should Democrats band together and start a polling firm that issues deliberately biased polls in their favor?
Since Polling is now at least 75% a PR function -the answer is a rather obvious YES .
There is no point for Democrats to do this, because the MSM would not take such polls seriously. Then why do they take the Republican-biased ones seriously? Ahhh, gee I wonder. The media loves Rasmussen nd Gallup, kind of like in that old song: “I’d rather go on hearing your lies, Than go on living without you.”
Way back when, Mike Royko had the best answer of all: lie.
As recounted by Vegas columnist Sherman Frederick:
I’ve believed for a long time that election polling is a net impediment to democracy. Without it, pols would be forced to work with issues instead of pandering to the polls. Why do we keep giving away our opinions for free so some cheats like Rasmussen can make money off of them? I wonder how many dedicated poll liars it would take to render polling useless. Sure would be worth finding out.