Jim Lindgren crunches the numbers and finds which pollsters were the most accurate in 2004. Rasmussen and Survey USA came out on top, with the worst being (unsurprisingly) the LA Times poll and Fox News’ polls also being wildly inaccurate (ironically Fox’s polls significantly overstated support for Kerry).
The lessons of the 2004 elections: as I’ve said before, partisan ID can change over time. Pollsters like Zogby who applied the partisan balance from 2000 were far off the mark while pollsters who either assumed an even partisan balance caught the increase in GOP registrations that erased the traditional Democratic advantage in voter ID.
Ironically, my estimates well understated Bush’s popular vote performance (I guessed around 49.5% at most and Bush got 51%), and overstated the Electoral College (Wisconsin, Minnesota, and Hawaii all went to Kerry). Indeed, from the analysis of the numbers, it appears that the 9/11 Democrats (not moral values voters) were enough to lift Bush’s numbers. It was terrorism, and not gay marriage that reelected President Bush. This was indeed a realigning election, and unless the Democrats start to understand Middle America rather than deride their faith and work ethic, a period of lasting Republican control similar to the Democratic coalition of FDR is quite possible. The real lesson of this election is that the Democrats drifted dangerously far out of the mainstream of American politics – and given the petulant reaction we’ve seen so far, it’s seems they’re failing to learn from their mistakes.
Was it that the Democrats drifted or that the rhetoric of terror and fear were simply a better sell that the rhetoric of hope? If the second thought is more correct then the first, why did New York, the prime target of 9/11, go Kerry instead of Bush? Is is, perhaps, that the residents of New York are ready for the rhetoric to change and policies to change as well?