What would be the reaction if, by election night, polls show a close race but slight advantage Obama, and in the end, the map looks like this?
(Image omitted)
Romney: 291
Obama: 247
Would there have been on overestimation of Obama support? Suppressed black turnout?
Such a situation would not surprise me at all. If the election turns out to favor Romney despite what the polls say, it's most likely the pollsters that are wrong. The explanation is fairly obvious: Democratic oversampling. The key to accurate polling is to have a sample representative of the people who actually vote on Election Day. I'm not aware of the specifics, but I imagine most pollsters assume a relatively even balance between Democrats and Republicans in their samples, at best. Some might give more weight to the Democrats. I remember a poll in July that had Obama in the lead, but the sample was D +19 (meaning the sample was weighted heavily towards the Democrats), which was denounced as ludicrous.
If Romney wins, I don't doubt the Democrats will let loose with all kinds of accusations, but the polls are far from totally reliable, even under the best of circumstances. My own predictions are based largely on the CU economic analysis and the results of previous presidential elections; I didn't even take poll results into account when I made my predictions about where the 9 "swing states" would lean.
