The strategic insight of the Bush re-election campaign in 2004 was that times had changed. The U.S. was divided 50-50 between the parties. The number of committed partisans had increased, and the number of true swing voters — as opposed to voters who say they are independent but reliably vote for one party — had shrunk. In this newly polarized country, no president could hope to achieve high approval ratings for very long.

It followed, though, that a president could win re-election even with approval ratings that would once have spelled doom. In a 50-50 America, every presidential election was a choice between the incumbent and the challenger and not just a referendum on the incumbent. If voters who didn’t approve of the incumbent could be persuaded to prefer him to the challenger, the incumbent would win…

His strategy will therefore be to make it a choice election. He is going to want the small number of swing voters to think: No, I’m not satisfied with how things are going and I have my doubts about Obama, but I’m more worried about the radicalism of the Republicans on Medicare and their fealty to big business.

As far as anyone can tell today, perceptions of the economy on Election Day are going to be closer to what they were in 2008 than what they were in 2004. That’s what the Republican referendum theory has going for it. But the president’s approval rating tells a different story. Since Obama’s honeymoon ended, it has moved in a fairly narrow range, never going below 44 percent and rarely going above 51 percent in the RCP average. It doesn’t put him in cinch-to-win or sure-loser territory.