Scott Rasmussen, president of Rasmussen Reports LLC, a national polling firm, is an active conservative author, speaker and columnist. The GOP relied heavily on Scott's company to produce polling numbers during the election season. Scott's polls became daily headlines on conservative websites like The Drudge Report. Fox News outsourced all their polling to Rasmussen. His polls were used repeatedly by conservative media types such as Rush Limbaugh, Ann Coulter and Laura Ingraham. Scott was the darling of conservatives, an advocate who appeared often on Fox and in conservative print media.
It was Rasmussen's poll numbers that helped give the illusion of widespread momentum toward Republicans in the weeks leading up to the election. After the first debate, all of a sudden, it was 2004 all over again. Rasmussen Reports showed Romney surging to the lead in swing states. Republicans had the big "Mo" - Momentum. Except, they didn't. On election night, it wasn't even close. Democrats won the presidency and gained seats in the Senate and House. And, GOP political pundits had few explanations for what happened. Scott Rasmussen's polls turned out to be a total fraud - so badly wrong that it begs the question - how could a professional pollster be THAT wrong?
In Senate races, Rasmussen had Missouri candidate Todd Akin (R) closing the gap on incumbent Claire McCaskill (D). Todd Akin, if you remember, was the candidate who uttered the phrase "legitimate rape." After Rasmussen's polling supported Akin's internal polling that showed him down by only 3, GOP super-pac money poured into Missouri. In the final week, Newt Gingrich toured the state with Akin. McCaskill destroyed Akin by 16 points.
In Florida, Rasmussen had incumbent Bill Nelson (D) ahead of Connie Mack (R) by only 3, within the margin of error. Nelson won by 13, well beyond the margin of error. Rasmussen had the Wisconsin Senate race tied. Tammy Baldwin (D) beat Tommy Thompson (R) by 6. In Connecticut, Rasmussen had Murphy (D) winning by 6 over McMahon (R). Murphy won by 12. In Ohio, Rasmussen showed Brown (D) and Mandel (R) tied; Brown won by 5. And, Rasmussen's presidential polling was just as inaccurate.
The weekend before the election, Rasmussen had Romney winning Florida, Virginia, Colorado and Iowa, with Wisconsin and Ohio tied. Obama won all of them, many beyond Rasmussen's margin of error. For instance, Rasmussen had Romney winning Colorado by 3, yet Obama won by 5. In Rasmussen's presidential popular vote tracking poll, Romney led Obama twelve straight days before the election, and up by an average of 6 points in swing states. So, how could a professional pollster get the election so wrong?
Rasmussen polled only landlines, a huge mistake. Most polling organizations include a mix of cellphones because they skew toward younger voters while landlines over represent older voters. Rasmussen's explanation was that he included results from online polling. That is grossly inadequate - online polling represents a different methodology and there's little to prevent a landline voter from being polled twice.
Since many Tea Party members identify themselves as independents, Rasmussen's polls over represented Romney's strength among true, unaffiliated independents. When Rasmussen compared his sample data with past voting histories of districts, counties, and states, he came up with overall totals that consistently skewed Republican.
Rasmussen was unable to identify demographic trends that altered historic voting patterns. White voters, as a percentage of the total vote, dropped from 77% in 2004 to 72% in 2012. Rasmussen's robocalls could not keep people on the phone long enough to ask follow up questions. It's not as though Rasmussen's sample size was so large, polling couldn't have been conducted by real people. Rasmussen's samples were as small as 500.
Yet, in spite of producing different results than most polling organizations, Republicans believed Rasmussen's polls to be accurate. Republicans have this perception, promulgated by conservative talk show hosts, commentators and libertarian think tanks, that there is a vast liberal conspiracy at work in the country. It's a conspiracy that includes the mainstream media, polling organizations, even going so far as to say educational institutions brainwash college students to be liberal. In other words, according to Republicans, the other polls were biased toward Democrats.
It is clear the composition of the American electorate is changing. On issues such as immigration and continued support for social programs, voters endorsed the Democratic point of view. There is irrefutable evidence that the GOP must modify it's positions on key issues and address demographic changes or risk continued losses in national elections. One would expect a polling organization to identify these trends well in advance. In the case of Scott Rasmussen and his company, Rasmussen Reports LLC, apparently not.
After being so wrong, if I were Scott Rasmussen, I'd consider another line of work. Oh, I forgot - Scott doesn't need a job; he's rich. He's a co-founder of the ESPN Sports Network. Why is a sports network guy doing political polling? I have no idea.
