Monday, November 29, 2010

Evidence Rasmussen polls are phony

Look at the contested Senate races from the 2010 election:

These are the races where at least one independent poll had the Democrat in the lead within the final month of the campaign. In addition to the closely watched Wisconsin, Delaware, California, and Connecticut races.

Final Rasmussen poll Results Bias
Wisconsin- Johnson +7 Johnson +5 Republican +2
West Virginia- Manchin +4 Machin +10 Republican +6
Washington- Rossi +1 Murray +4 Republican +5
Pennsylvania- Toomey +4 Toomey +2 Republican +2
Nevada- Angle +4 Reid + 5 Republican +9
Illinois- Kirk +4 Kirk +2 Republican +2
Delaware- Coons +11 Coons +17 Republican +6
Connecticut - Blumenthal +7 Blumenthal +12 Republican +5
Colorado - Buck +4 Bennet +1 Republican +5
California - Boxier +3 Boxer +10 Republican +7

It could be argued that the Kentucky race should be on here:
Kenucky - Paul +12 Paul +12 no bias

... but the race had clearly broken for the Republican.

Here's a few other non-competitive races:

Final Rasmussen poll bias

Vermont Republican + 2
Utah Republican +4
South Carolinaa Republican +3
Oregon Republican + 7

The facts show that the Rasmussenreports methodology was flawed. Or that the results were faked. I would lean towards flawed systemic problems because it is so easy to manipulate them. Why fake results when you can set it up so that you get what you want in a more legitimate way.

Democrats should file campaign finance violations against Scott Rasmussen because he was clearly donating to the Republicans without disclosure and financial limits.

Think how much it was worth, Candidate Fiorina or O'donnell to say, "I'm closing in on the Democrat, look at the latest indepedent Rasmussen poll".

I remember when the Connecticut poll came out. Given what had happened in the primary where Christine O'donnell closed quickly on Rep. Castle, the Rasmussen poll was evidence of a possible media backlash. Fox News, Republicans, and those on her side, now had ammunition to rally their base. It didn't work, but it could have.

Likewise with the healthcare polls, and all the other issue surveys that rasmussen releases.

They should be forced to declare as Republicans or "Conservative pollster" just like PPP is always labeled Democratic pollsters.

No comments: