The Price Of Tea In China

Some people have been passing around this piece from Conn Carroll (lot of double letters there) that says "Gallup state numbers predict huge Obama loss." I don't know if Carroll wrote his own headline, but there's a disconnect between it and the body of his piece.

First, of course, these numbers predict nothing.  They're just aggregated data from 2011, and things can easily change by election day.  In fact, they already have.  Last year, Obama averaged 44% approval.  His approval rating is higher at present, so he's already moving back up.

Second, these numbers only show his relative popularity in the 50 states.  What matters is his relative popularity against his opponent.  Even if Obama averaged 30% popularity, if the Republican averaged 20% popularity, Obama would probably beat him.

Even in the weak approval year of 2011, Obama was positioned to do well in states with 159 electoral votes, and to do poorly in states with 153 electoral votes.  The election will probably be determined by 12 swing states, especially highly populated ones such as Ohio, Florida and Pennsylvania.  Yes, it's true, if he's still stuck around 45% approval in such states, he could be in trouble (assuming the Republican candidate doesn't turn people off even more), but he doesn't even need to get above 50% to win.  Bush was reelected in 2004 with only 48% approval.