Nate Silver has an exhaustive and exhausting post on how Republicans could win the White House while losing further slices of the Latino vote. It's doable, but extremely hard. They have to hope for a double-dip recession and a divisive appeal to white voters. That's the kind of short-term idea that leads to long-term defeat (i.e. Karl Rove might love it). I didn't realize this:

In 2008, the Latino vote made the difference in the outcome of three states: New Mexico, where about 2 in 5 voters identify as Hispanic, as well as -- somewhat surprisingly -- Indiana and North Carolina -- where Obama lost nonhispanic voters by a tiny margin and was put over the top by Hispanic votes. It probably also made the difference, believe it or not, in the 2nd Congressional District of Nebraska -- Omaha actually has a decent-sized Hispanic minority -- although the exit polls aren't detailed enough to let us know for sure.

Nate's bottom line:

This is the sort of electoral future the GOP might have to contemplate if they start losing the Hispanic vote by margins of 3:1, 4:1 or more. Giving up on New Mexico, Nevada and Colorado is a feasible, and perhaps even wise, strategy. But if they don't thread the needle just perfectly, and they make it difficult for themselves to win back Florida, while putting Arizona and perhaps even Texas increasingly into play, their task will become nearly impossible.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.