Jonah’s Crystal Ball and Rasmussen’s Turnout Model
Don’t miss Jonah’s look back on Obama’s term, but then get back to turning out the GOP vote.
Scott Rasmussen is a superb pollster and he has Obama’s 5 point lead outside of the margin of error, but he also employs a prediction model that assumes a 6.5 point Democrat turnout advantage. Here are Rasmussen’s turnout assumptions:
For polling data released during the final days of Election 2008, the partisan weighting targets will be 39.9% Democratic, 33.4% Republican, and 26.7% unaffiliated. For the preceding week, the partisan weighting targets were 40.0% Democratic, 32.8% Republican, and 27.2% unaffiliated.
It was a 2% gap in 2004. Lots of pundits are predicting a demoralized GOP base and subsequent low turnout in support of Rasmussen’s and others predictions, but having just returned from Colorado, Minnesota and Ohio where first-hand experiences with volunteers and rallies conveyed exactly the opposite, I think the potential for surprise on the turnout gap is great.
Any such surprises will be put down to the mythical “Bradley effect,” when in fact any gap between the polling and the result will be traceable right to the turnout assumptions.