Nate Silver just tweeted: “FWIW, there’s about a 4 pt difference between generic ballot polls that included cellphones and those which didn’t.”
That’s a huge spread by methodology, though conceivably Rasmussen’s ‘house effect’ may be skewing the difference we can attribute to methodology. Regardless, it’s looking more and more like 2010 may be the last year when pollsters were able to write off the cell phone issue.
