“On the Republican side, this was the worst cycle ever for polling and there’s nothing that even comes close to it,” said GOP strategist Curt Anderson, who helms the media and polling firm OnMessage. “It was a colossal disaster and it wasn’t confined to the presidential campaign.”
Anderson said the proliferation of different groups — campaigns, outside spenders, etc. — polling the same states had made it exceptionally difficult to decode the political state of play and develop strategy accordingly.
“It seemed like you had people taking polls in different universes at the same time,” Anderson said. “Three, four points [of disagreement], that’s the margin of error, but I’m talking 10-, 15-point differences.”
It’s not as if the party is totally at a loss to explain where it went off course. Sources familiar with Romney’s polling say that it underestimated the Democrats’ 6-point voter identification edge, nationally, and put far too much stock in what one Republican operative called “false signs of Republican enthusiasm.” Multiple Republican pollsters also acknowledged that they misjudged how many young people and minorities would show up to vote.
“We need to rethink what voters we screen out, because clearly, we’re screening people out who are going to vote, and that’s manifestly affecting the numbers we’re looking at,” said NRSC executive director Rob Jesmer, who said the presidential race effectively swamped the GOP “likely voter” model. “It’s more apt in a presidential cycle, but we need to think about it.”
One GOP strategist involved in congressional races said the party’s pollsters need to gather “a conference of sorts between all of them to figure out what to do going forward,” pointing especially to the question of how to sample cell phone users who make up a growing share of the electorate.