Lynn Vavreck, a professor at UCLA, postulates that the mid-terms won't come down to swing voters. She provides the first data I've ever seen on how 2008 Obama and McCain voters voted in 2010, showing that only 6% of the voters swung to the other party in the congressional election. I appreciate that information and it answers the question of whether 2008 Obama voters flipped. A few thoughts:
1. She compares the 2008 Presidential vote to the 2010 congressional vote. The inherent problem with this is that some Obama voters voted Republican for congress in 2008 and vice-versa. Obama won 8 California districts where the Republican congressman also won. For instance, Obama took 51% of the vote in CA-25, but the Democratic candidate took only 42%. That could account for all the "switchers." Or not. So who did the 2010 voters vote for congress in 2008?
2. She implies that the 2008 Obama/McCain voters who didn't show up were loyal Democratic or Republican voters. Some might be. Others might be swing voters. Even those who aren't swing voters are likely those who never vote in anything other than a Presidential election. They aren't loyal to the party, so you'll have a tough time turning them out in a mid-term.
3. She doesn't address the voters who voted in 2010 and not 2008. While this isn't an overwhelming number of voters, they could be swing voters.
4. Just because around 6% of voters actually swung doesn't mean that there are only 6% swing voters. They may have voted the same party in 2008 and 2010, but that doesn't mean they did so before 2008 or did in 2012. The parties are able to keep some swing voters.
Despite my concerns, I consider this an enlightening insightful article.