Wednesday, November 3, 2010

Rasmussen Reports polls widely wrong

The Clarion Content, among many others, has questioned the accuracy and neutrality of polls from the Rasmussen Reports. They were founded by a former employee of the coronation, errr, campaign, of King George the II. They consistently phrase their polling questions with just a little twist to produce the desired result.

Their defense? They were right on the numbers in 2004 and 2006.

Last night, however, they were outed. According to our friend Nate Silver, over at the 538, "Rasmussen polls quite consistently turned out to overstate the standing of Republicans tonight. Of the roughly 100 polls released by Rasmussen or its subsidiary Pulse Opinion Research in the final 21 days of the campaign, roughly 70 to 75 percent overestimated the performance of Republican candidates, and on average they were biased against Democrats by 3 to 4 points."

Interestingly, it would be our argument that overstating your candidates odds at the polls actually hurts their chances. It would be analagous to the brass continuously telling American troops in Vietnam that the enemy was weak, under-armed, on the verge of starvation and primed for defeat. These lies and exaggerations actually harmed United States soldiers and Marines by leading them to regularly underestimate the enemy. We would argue that the Rasmussen Reports did something of the same for right-wing candidates last night.

No comments: