POLITICS AUGUST 22, 2013
Scott Rasmussen has left Rasmussen Reports, one of the least accurate pollsters of the last two elections. From that perspective, you might expect that Rasmussen was fired for bad surveys. His polls were biased toward Republican candidates in two consecutive cycles, outrageously so in 2012. He refused to interview voters with a cell phone, even though mounting research confirms they tilt toward Democratic leaning groups. He weighted his samples to a fantasy electorate where there are millions more white, Republican voters. And unlike Gallup, which is in the middle of an extensive post-election effort to improve its polling, Rasmussen's post-election rethinking involved reweighting their tracking poll to the 2012 exit polls, even though you really, really can't do that. Really embarassing stuff.
But according to the Rasmussen Reports press release, Rasmussen left because of disagreements over business strategy. The release says that Rasmussen’s “methodologies and protocols… continue to guide and inform the company’s public opinion survey techniques.” They even asserted that Rasmussen’s methodology is “widely acknowledged as among the most accurate and reliable in the industry.” Unbelievable.
So there’s not yet evidence that post-Rasmussen Rasmussen Reports won’t tolerate the same flawed polls that adulterated the polling averages for most of 2012. For now, Rasmussen Reports looks like it's still as bad as Rasmussen, even without Rasmussen.