2012 Prediction

Thanks to the calculator at USElectionAtlas (and borrowing liberally from polling data on Wikipedia), here is my prediction for the 2012 presidential election, assuming a two-way contest between Barack Obama and Mitt Romney:

Romney takes Florida [29], Maine’s second congressional district [1], New Hampshire [4], Virginia [13]; Obama takes Arizona [11]; Nevada [6] is a toss-up (the only poll taken there in the last few months has Obama and Romney tied). Obama leads in enough swing states to hang on to the White House, and takes 317 electoral votes (323 if Nevada goes his way) – down from his tally of 365 in 2008, but still well over the 270 needed.

All other potential Republican candidates fare much worse than Romney in state-by-state polls against Obama, so it’s safe to assume an Obama victory by a larger margin if the GOP settles on a non-Romney nominee. There is, of course, the specter of a Michael Bloomberg independent candidacy. For this, I made an educated guess that Bloomberg could win New York, New Jersey, Connecticut, California, and Florida, and that he would eat into Romney’s numbers elsewhere, delivering Obama a few more states (Alaska, Georgia, Maine’s second district, Montana, New Hampshire, Virginia, as well as tipping Nevada into the Democratic column).

The result: Obama 258, Romney 146, Bloomberg 134, so the election would go to the House of Representatives.