I don't think I've had one post in three months that wasn't on California. So I'll take a break and highlight this Democracy Corps poll. Democracy Corps is a Democratic polling firm where James Carville and Stanley Greenberg are the principles.You'd expect Democratic bias. That's not the case.
The poll is of likely voters, not the registered voters that most polls do. So this should be a better predictor of the actual electorate. Obama leads Romney 49%-46%, while Republicans lead by 1 point on the generic ballot. That sound pretty good for Obama. It isn't.
The poll is 37%D/31%R/32%I. Democrats only had a lead of 4+ points advantage once, in 2008. Every Presidential year from 1984 to 2000 had D+2-4 and 2004 was even. So not only would it be the second most Democratic electorate, but it'd be the lowest percentage Republican turnout, as Republicans were 32% of the electorate in 2008. This'd imply that Republicans are dispirited and disinterested and about as likely to vote as they were in 2008.
I don't think anyone paying attention would think this year will be anything like 2008 for Republicans. It's likely going to be closer to 2010. Democrats will likely be closer to 2008 than 2010, so I'd expect something like 38%D/36%R/26% I.
They don't have in tabs, but these would work.
Generic D 83%/8%/37%
Generic R 11%/87%/47%
That's an awfully big lead with independents, far bigger than anything suggests. If we change the poll to the breakdown above we come up with a 48%-48% Obama-Romney tie and a 48%-44% Republican lead on the generic ballot. That isn't a big shift, but if Obama wins the popular vote by 3 points he wins the Presidency. If he gets a similar number of votes to Romney, then he might not.