As of October 10th, 2008.
Obama: 343
McCain: 184
Tie: 11
- North Carolina went from a tie to slightly McCain.
- Missouri went from slightly Obama to a tie.
The Webmaster at electoral-vote.com has this to say about the accuracy of polling this year:
Polling's Perfect Storm?
According to Mark Blumenthal at National Journal, there are three factors at work that may cause polls to be way off this year. They are:
- 20% of all households don't have a landline and they are mostly younger voters
- The "Bradley Effect" in which white voters tell pollsters they will vote for a black guy but don't
- Likely voter models may not capture the electorate correctly
Cell-phone-only voters are young and strongly pro-Obama and may cause his strength to be understated. The effect can be compensated for by weighting 18-29 year olds in sufficient numbers (but see below). The second effect is unknown although it did not appear to be a big factor in the primaries. This effect would cause McCain's strength to appear weaker than it really is. Finally, the likely voter models assume there is some way to tell who will vote and who won't. This year that is very hard to gauge and demographic models based on the 2004 exit polls may not be worth much. On the other hand, these effects may all cancel. It is a tough call.
