Amy Walter: “This, of course, is the billion-dollar question. And, it is one that pollsters and campaigns spend all their time and energy trying to get right. We are seeing an unprecedented level of transparency this year by pollsters and other media outlets about how they arrived at their results. Survey Monkey’sMark Blumenthal explained how they arrived at their ‘likely voter’ model. The New York Time’s Upshot farmed their raw polling data from a Florida poll to four different pollsters to show how each of them came up with a different margin for Clinton/Trump.”
“Perhaps the best new tool on the market was designed by friend of the Cook Political Report and political scientist Ken Goldstein which goes beyond just “averaging” the various polls to looking at the assumptions the pollsters made to get to the final result.”
“At the end of the day, no one knows exactly what turn-out will be until people actually turn-out. However, we know a couple things: first, only one campaign has a data-driven turn-out program (Clinton) to help identify and get their voters to the polls; and second, Democrats have a demographic advantage, but only if those voters – namely non-white and younger, turn out. The Obama coalition has helped to elect only one Democrat – Barack Obama. In the other two mid-term elections they have stayed home. Trump, meanwhile, needs to hit an unprecedented level of support from white voters and/or hope for a lower turnout among nonwhites.”
